Docstoc

awprftr

Document Sample
awprftr Powered By Docstoc
					                                AWPR
  Automatic Wireless Phone Registration
                               Table of Contents


                                      Walt Magnussen, Ph. D.
                                             Project Sponsor




                                           Pierre Catala
                                           Dr. Jay Porter
                                             Faculty Advisors




                                     Ringer Communications



                                         December 5, 2005



                                                                Team Members:
605 Highland Street
                                                                    John Kelley
College Station, TX, 77840                                            Chris Rue
(512) 639-4284                                                     Kara Flanary
ringercomm@gmail.com
www.ringercommunications.com                                     Ranesha Cobb
                                                                                                                                             December 5, 2005




                                                                 Table of Contents

I. PROJECT OVERVIEW ....................................................................................................................................... 1
    A. GENERAL ............................................................................................................................................................. 1
    B. BACKGROUND ...................................................................................................................................................... 2
    C. TECHNICAL CHALLENGE ...................................................................................................................................... 2
    D. BENEFITS ............................................................................................................................................................. 4
    E. TECHNICAL REPORT STRUCTURE ......................................................................................................................... 4
II. PROJECT SCOPE ............................................................................................................................................... 6
III. FUNCTIONAL REQUIREMENTS .................................................................................................................. 8
    A.   SYSTEM INTEGRATION ......................................................................................................................................... 8
    B.   ANALYSIS .......................................................................................................................................................... 12
    C.   TESTING ............................................................................................................................................................. 13
    D.   ADDITIONAL DOCUMENTATION ........................................................................................................................ 14
IV. MANAGEMENT............................................................................................................................................... 15
    A.   MILESTONES ..................................................................................................................................................... 15
    B.   DELIVERABLES .................................................................................................................................................. 17
    C.   EARNED VALUE PROCESS .................................................................................................................................. 20
    D.   INDIVIDUAL CONTRIBUTIONS ............................................................................................................................ 22
V. TECHNICAL NETWORKING ........................................................................................................................ 24
    A.   BASIC SIP CALL FLOW ...................................................................................................................................... 25
    B.   PROJECT PROTOCOLS......................................................................................................................................... 26
    C.   CALL REGISTRATION ......................................................................................................................................... 28
    D.   CALL SCENARIOS .............................................................................................................................................. 29
VI. EQUIPMENT .................................................................................................................................................... 35
    A. NETWORK EQUIPMENT ...................................................................................................................................... 35
      1. Nortel 2230 Access Port............................................................................................................................... 35
      2. Nortel BayStack 460 Switch ......................................................................................................................... 36
      3. Nortel WLAN Security Switch 2270 ............................................................................................................. 36
      4. HP iPAQ 6315 ............................................................................................................................................. 36
    B. EQUIPMENT CONFIGURATIONS .......................................................................................................................... 37
      1. Nortel BayStack 460 Switch ......................................................................................................................... 37
      2. Nortel WLAN Security Switch 2270 ............................................................................................................. 43
      3. HP iPAQ 6315 ............................................................................................................................................. 50
      4. PSTN/IP Gateway ........................................................................................................................................ 62
    C. TESTING EQUIPMENT ......................................................................................................................................... 64
      1. IXIA 1600 Traffic Generator........................................................................................................................ 64
      2. Berkeley Varitronics “Grasshopper” 802.11 Tester.................................................................................... 64
      3. Agilent Voice Quality Tester ........................................................................................................................ 65
      4. IXIA IxWLAN ............................................................................................................................................... 65
      5. Linksys WET Wireless Bridge ...................................................................................................................... 65
VII. SOFTWARE .................................................................................................................................................... 66
    A. WIRELESS VALLEY’S “SITE PLANNER” ............................................................................................................. 66
    B. CAMTASIA STUDIO 3 ......................................................................................................................................... 66
    C. PCTEL ROAMING CLIENT ................................................................................................................................. 67



           AWPR – Automatic Wireless Phone Registration
                                                                                                                                            December 5, 2005


    D. POCKET PC CONTROLLER ................................................................................................................................. 67
    E. IXIA IXEXPLORER ............................................................................................................................................. 67
    F. PASSMARK WIRLESSMON .................................................................................................................................. 68
VIII. ANTENNA SYSTEM ANALYSIS ............................................................................................................... 69
    A. SOFTWARE VALIDATION ................................................................................................................................... 69
      1. Introduction.................................................................................................................................................. 69
      2. Procedure..................................................................................................................................................... 70
      3. 802.11 Propagation Test.............................................................................................................................. 80
    B. ANTENNA SELECTION ........................................................................................................................................ 84
      1. Wi-Sys Communications 802.11b/g Indoor/Outdoor Waveguide................................................................. 85
      2. Trendnet 6dBi Indoor Directional Patch Antenna ....................................................................................... 89
IX. TESTING ........................................................................................................................................................... 92
    A. TESTING OVERVIEW .......................................................................................................................................... 92
    B. NETWORK ASSOCIATION TIMING ...................................................................................................................... 92
    C. AUTOMATIC CALL REGISTRATION SIGNAL LEVEL ............................................................................................ 94
    D. VOICE QUALITY ................................................................................................................................................ 95
      1. Various Received Signal Strengths .............................................................................................................. 95
      2. Wired Traffic Loading.................................................................................................................................. 96
      3. Wireless Traffic Loading............................................................................................................................. 96
    E. BATTERY LIFE .................................................................................................................................................. 97
    C. ANTENNA TESTING ............................................................................................................................................ 98
X. CONCLUSIONS ............................................................................................................................................... 106
XI. RECOMMENDATIONS FOR FURTHER RESEARCH ........................................................................... 108
XII. GLOSSARY ................................................................................................................................................... 111
XIII. APPENDIX .................................................................................................................................................. 113
    A. NETWORK ASSOCIATION TIMING TEST ................................................................................................................ I
       1. Objective .......................................................................................................................................................... i
       2. Equipment ...................................................................................................................................................... ii
       3. Test Plan ........................................................................................................................................................ ii
       4. Test Results ................................................................................................................................................... vi
    B. AUTOMATIC CALL REGISTRATION SIGNAL LEVEL ............................................................................................. IX
       1. Objective ....................................................................................................................................................... ix
       2. Equipment ...................................................................................................................................................... x
       3. Test Plan ........................................................................................................................................................ x
       4. Test Results ................................................................................................................................................... xi
    C. VOICE QUALITY ............................................................................................................................................... XIII
       1. Various Received Signal Strength Levels.................................................................................................... xiii
       2. Wired Traffic Loading................................................................................................................................. xxv
       3. Wireless Traffic Loading.............................................................................................................................. xli
    D. BATTERY LIFE................................................................................................................................................. LXIX
       1. Objective .................................................................................................................................................... lxix
       2. Equipment .................................................................................................................................................. lxix
       3 GSM Testing Procedure................................................................................................................................lxx
       4. Testing Procedure for 802.11B ................................................................................................................ lxxiii
       5 GSM Test Results .......................................................................................................................................lxxiv
       6 802.11B Test Results ..................................................................................................................................lxxvi
       7 Duty Cycle................................................................................................................................................lxxviii




           AWPR – Automatic Wireless Phone Registration
                                                                                                                                     December 5, 2005


                                                                 List of Figures
FIGURE 1. NETWORK BLOCK DIAGRAM. ........................................................................................................................ 8
FIGURE 2. AVERAGE CLIENT THROUGHPUT VS. SIGNAL STRENGTH.............................................................................. 11
FIGURE 3. MILESTONES. .............................................................................................................................................. 15
FIGURE 4. DELIVERABLES FOR SEPTEMBER AND OCTOBER. ........................................................................................ 17
FIGURE 5. DELIVERABLES FOR NOVEMBER AND DECEMBER........................................................................................ 19
FIGURE 6. EVP (SPI VS. CPI). ...................................................................................................................................... 22
FIGURE 7. FUNCTIONAL DESIGN................................................................................................................................... 24
FIGURE 8. SIP CALL SETUP\BREAKDOWN. ................................................................................................................... 25
FIGURE 9. OSI MODEL COMPARED TO PROJECT PROTOCOL STACK. ............................................................................ 26
FIGURE 10. PSTN TO GSM CALL SCENARIO................................................................................................................ 30
FIGURE 11. VOIP TO GSM CALL SCENARIO................................................................................................................. 31
FIGURE 12. VOIP TO VOIP CALL SCENARIO................................................................................................................. 32
FIGURE 13. PSTN TO VOIP CALL SCENARIO................................................................................................................ 33
FIGURE 14. EQUIPMENT AT ITEC LAB. ........................................................................................................................ 35
FIGURE 15. BAYSTACK 460 WEB INTERFACE............................................................................................................... 37
FIGURE 16. POE GLOBAL POWER MANAGEMENT......................................................................................................... 38
FIGURE 17. POE PORT POWER MANAGEMENT. ............................................................................................................ 39
FIGURE 18. BAYSTACK 460 VLAN CONFIGURATION. ................................................................................................. 40
FIGURE 19. BAYSTACK PORT-BASED VLAN CREATION.............................................................................................. 41
FIGURE 20. BAYSTACK VLAN PORT ASSOCIATION..................................................................................................... 42
FIGURE 21. BAYSTACK VLAN PORT CONFIGURATION................................................................................................ 43
FIGURE 22. NORTEL SECURITY SWITCH LOGIN. ........................................................................................................... 44
FIGURE 23. SECURITY SWITCH WLAN CONFIGURATION SCREEN................................................................................ 44
FIGURE 24. SECURITY SWITCH NEW WLAN SCREEN. ................................................................................................. 45
FIGURE 25. SECURITY SWITCH PORT CONFIGURATION. ............................................................................................... 46
FIGURE 26. SECURITY SWITCH WIRELESS SETTINGS.................................................................................................... 47
FIGURE 27. SECURITY SWITCH ACCESS POINT CONFIGURATION.................................................................................. 47
FIGURE 28. ACCESS POINT RADIO CONFIGURATION..................................................................................................... 48
FIGURE 29. SECURITY SWITCH SUPPORTED DATA RATES. ........................................................................................... 49
FIGURE 30. HP IPAQ START MENU.............................................................................................................................. 50
FIGURE 31. HP IPAQ LAUNCHING FILE EXPLORER...................................................................................................... 51
FIGURE 32. HP IPAQ EXECUTING THE VERISIGN SOFT CLIENT INSTALLER................................................................. 52
FIGURE 33. HP IPAQ LAUNCHING THE UNIVERSAL PHONE. ........................................................................................ 53
FIGURE 34. HP IPAQ UNIVERSAL PHONE IN GSM MODE. ........................................................................................... 53
FIGURE 35. HP IPAQ UNIVERSAL PHONE WI-FI CONFIGURATION............................................................................... 54
FIGURE 36. HP IPAQ AWPR WIRELESS NETWORK GENERAL SETTINGS..................................................................... 55
FIGURE 37. HP IPAQ AWPR WIRELESS NETWORK SECURITY SETTINGS. ................................................................... 56
FIGURE 38. HP IPAQ CONNECTING TO AN AWPR WIRELESS NETWORK..................................................................... 56
FIGURE 39. HP IPAQ SUCCESSFUL CONNECTION TO AWPR WIRELESS NETWORK. .................................................... 57
FIGURE 40. HP IPAQ UNIVERSAL PHONE ATTEMPTING TO CONNECT TO VERISIGN'S SIP SERVER. ............................ 58
FIGURE 41. HP IPAQ UNIVERSAL PHONE SIP ERROR MESSAGE.................................................................................. 58
FIGURE 42. HP IPAQ STARTING THE SIP CONFIGURATION.......................................................................................... 59
FIGURE 43. HP IPAQ UNIVERSAL PHONE SIP CONFIGURATION................................................................................... 60
FIGURE 44. HP IPAQ UNIVERSAL PHONE SIP AUTHENTICATION SETTINGS. ............................................................... 61
FIGURE 45. HP IPAQ UNIVERSAL PHONE SUCCESSFUL VOIP REGISTRATION.............................................................. 62
FIGURE 46. SIMULATED ISOTROPIC RADIUS AT -50 DBM. ............................................................................................ 72
FIGURE 47. SIMULATED ISOTROPIC RADIUS AT -60 DBM. ............................................................................................ 73
FIGURE 48. SIMULATED ISOTROPIC RADIUS AT -70 DBM. ............................................................................................ 73
FIGURE 49. SIMULATED ISOTROPIC RADIUS AT -80 DBM. ............................................................................................ 74
FIGURE 50. SIMULATED ISOTROPIC RADIUS AT -90 DBM ............................................................................................. 74
FIGURE 51. SIMULATED CARDIOID RANGE AT -50 DBM. .............................................................................................. 77
FIGURE 52. SIMULATED CARDIOID RANGE AT -60 DBM. .............................................................................................. 78



           AWPR – Automatic Wireless Phone Registration
                                                                                                                                      December 5, 2005


FIGURE 53. SIMULATED CARDIOID RANGE AT -70 DBM. .............................................................................................. 79
FIGURE 54. SIMULATED CARDIOID RANGE AT -80 DBM. .............................................................................................. 79
FIGURE 55. SIMULATED CARDIOID RANGE AT -90 DBM. .............................................................................................. 80
FIGURE 56. MODELED BASE STATIONS AND THEIR COVERAGE. .................................................................................. 81
FIGURE 57. EDIT DIALOG BOXES FOR PARTITIONS....................................................................................................... 82
FIGURE 58. STAR VISION'S PREDICTED RSSI VALUES.................................................................................................. 83
FIGURE 59. ADVANTGX'S PREDICTED RSSI VALUES................................................................................................... 83
FIGURE 60. WI-SYS 8DBI WAVEGUIDE ANTENNA........................................................................................................ 85
FIGURE 61. WAVEGUIDE HORIZONTAL RADIATION PATTERN. ..................................................................................... 85
FIGURE 62. WAVEGUIDE VERTICAL RADIATION PATTERN........................................................................................... 85
FIGURE 63. WAVEGUIDE ANTENNA LAYOUT ON FIRST FLOOR. ................................................................................... 87
FIGURE 64. WAVEGUIDE ANTENNA LAYOUT ON SECOND FLOOR. ............................................................................... 88
FIGURE 65. TRENDNET 6DBI PATCH ANTENNA. ........................................................................................................... 89
FIGURE 66 - PATCH HORIZONTAL PATTERN ................................................................................................................. 89
FIGURE 67 - PATCH VERTICAL PATTERN ...................................................................................................................... 89
FIGURE 68. PATCH ANTENNA LAYOUT ON FIRST FLOOR.............................................................................................. 90
FIGURE 69. PATCH ANTENNA LAYOUT ON SECOND FLOOR.......................................................................................... 91
FIGURE 70. MOS VS RECEIVED SIGNAL STRENGTH FOR PSTN & VOIP....................................................................... 95
FIGURE 71. MOS VS. TRAFFIC FOR PSTN & VOIP....................................................................................................... 96
FIGURE 72. CHANNEL 9 ACCESS POINT’S RADIATION ON FIRST FLOOR. .................................................................... 100
FIGURE 73. CHANNEL 9 ACCESS POINT’S RADIATION ON THE SECOND FLOOR. ......................................................... 101
FIGURE 74. CHANNEL 3 SVT ACCESS POINT’S RADIATION ON THE FIRST FLOOR. ..................................................... 102
FIGURE 75. CHANNEL 3 SVT ACCESS POINT’S RADIATION ON THE SECOND FLOOR.................................................. 103
FIGURE 76. CHANNEL 3 UPSTAIRS ACCESS POINT’S RADIATION ON THE FIRST FLOOR.............................................. 104
FIGURE 77. CHANNEL 3 UPSTAIRS ACCESS POINT’S RADIATION ON THE SECOND FLOOR.......................................... 105
FIGURE 78. SIP REGISTRATION PACKET CAPTURE. ....................................................................................................... IV
FIGURE 79. VIDEO CAPTURE OF PHONE IN VOIP MODE. ................................................................................................ V
FIGURE 80. VIDEO CAPTURE OF PHONE IN GSM MODE. ............................................................................................... VI
FIGURE 81. PSTN <-> VOIP TEST SETUP. .................................................................................................................... XV
FIGURE 82. PSTN <-> GSM TEST SETUP. ...................................................................................................................XVI
FIGURE 83. VOIP <-> VOIP TEST SETUP. ....................................................................................................................XVI
FIGURE 84. SIGNAL LOSS TEST...................................................................................................................................XVII
FIGURE 85. VOICE QUALITY TEST...............................................................................................................................XIX
FIGURE 86. SIGNAL STRENGTH LEVELS........................................................................................................................ XX
FIGURE 87. NORMAL DISTRIBUTION GRAPH. ..............................................................................................................XXI
FIGURE 88. MOS VS. RSSI (PSTN->VOIP AND VOIP->PSTN). ................................................................................XXII
FIGURE 89. VOIP VS. GSM VOICE QUALITY. ........................................................................................................... XXIII
FIGURE 90. VOIP TO VOIP VOICE QUALITY............................................................................................................. XXIV
FIGURE 91. WIRED TRAFFIC LOADING NETWORK. ................................................................................................... XXVI
FIGURE 92. IXEXPLORER READY STATE CHANGE. .................................................................................................. XXVII
FIGURE 93. LOGIN SCREEN...................................................................................................................................... XXVII
FIGURE 94. IXEXPLORER CHASSIS OWNERSHIP...................................................................................................... XXVIII
FIGURE 95. IXEXPLORER NEW TRAFFIC STREAM. .................................................................................................... XXIX
FIGURE 96. NAME TRAFFIC STREAM AND SET MAXIMUM RATE. .............................................................................. XXX
FIGURE 97. TRAFFIC STREAM'S FRAME DATA. ......................................................................................................... XXXI
FIGURE 98. IP HEADER............................................................................................................................................ XXXII
FIGURE 99. OPEN PROTOCOL WINDOW. ................................................................................................................. XXXIII
FIGURE 100. ADDING DEFAULT GATEWAY ADDRESS. ........................................................................................... XXXIV
FIGURE 101. NEW IP RANGE. ................................................................................................................................. XXXIV
FIGURE 102. FIELDS FOR NEW IP RANGE. ............................................................................................................... XXXV
FIGURE 103. ARP................................................................................................................................................... XXXVI
FIGURE 104. MAC ADDRESS RETRIEVED............................................................................................................... XXXVI
FIGURE 105. BEGIN TRAFFIC GENERATION. .......................................................................................................... XXXVII
FIGURE 106. VOICE QUALITY VS. TRAFFIC LOAD. ................................................................................................. XXXIX
FIGURE 107. WIRELESS BRIDGE TEST SETUP. ............................................................................................................ XLII



           AWPR – Automatic Wireless Phone Registration
                                                                                                                                       December 5, 2005


FIGURE 108. BRIDGE'S WEB INTERFACE................................................................................................................... XLIII
FIGURE 109. BASIC CONFIGURATION OF THE BRIDGE. .............................................................................................. XLIII
FIGURE 110. SITE SURVEY WINDOW. ........................................................................................................................ XLIV
FIGURE 111. SITE SURVEY CONFIRMATION................................................................................................................ XLV
FIGURE 112. WEP KEY SECURITY WARNING............................................................................................................. XLV
FIGURE 113. WEP KEY WINDOW.............................................................................................................................. XLVI
FIGURE 114. ADVANCED BRIDGE SETTINGS.............................................................................................................XLVII
FIGURE 115. STATUS WINDOW................................................................................................................................XLVIII
FIGURE 116. SWITCH COMMANDS...........................................................................................................................XLVIII
FIGURE 117. BRIDGE LINK TEST. .............................................................................................................................. XLIX
FIGURE 118. IXIA TRAFFIC STREAM SETUP................................................................................................................... L
FIGURE 119. IXIA FRAME DATA SETUP........................................................................................................................ LI
FIGURE 120. IXIA IP HEADER SETUP........................................................................................................................... LII
FIGURE 121. IXIA MAC ADDRESS SETUP................................................................................................................... LIII
FIGURE 122. IXIA PING TEST...................................................................................................................................... LIII
FIGURE 123. IXIA TRAFFIC GENERATOR STATS WINDOW.......................................................................................... LIV
FIGURE 124. FLOODED PING TEST................................................................................................................................ LV
FIGURE 125. LINK TEST FOR ANOTHER CLIENT. .......................................................................................................... LV
FIGURE 126 ETHEREAL WIRELESS CAPTURE............................................................................................................... LVI
FIGURE 127. IXWLAN TEST NETWORK. ..................................................................................................................... LVI
FIGURE 128. IXWLAN TELNET SESSION.................................................................................................................... LVII
FIGURE 129. IXWLAN LOGIN.................................................................................................................................... LVII
FIGURE 130. IXWLAN STARTUP WIZARD ................................................................................................................ LVIII
FIGURE 131. VSTA SCREEN. ....................................................................................................................................... LIX
FIGURE 132. NEW IXWLAN GROUP WINDOW (VSTA SETTINGS). .............................................................................. LX
FIGURE 133. NEW IXWLAN GROUP WINDOW (TRAFFIC SETTINGS)........................................................................... LXI
FIGURE 134. NEW IXWLAN GROUP WINDOW (SECURITY SETTINGS). ...................................................................... LXII
FIGURE 135. JOIN SUT. ............................................................................................................................................. LXIII
FIGURE 136. RUNNING THE VIRTUAL STATION. ........................................................................................................ LXIII
FIGURE 137. IXIA IP ADDRESS SETUP. ..................................................................................................................... LXIV
FIGURE 138. IXIA MAC ADDRESS SETUP.................................................................................................................. LXV
FIGURE 139. NEW MONITOR. ..................................................................................................................................... LXV
FIGURE 140. NEW MONITOR WINDOW. ..................................................................................................................... LXVI
FIGURE 141. MONITORS DISPLAY. ...........................................................................................................................LXVII
FIGURE 142. IXWLAN ETHEREAL CAPTURE............................................................................................................LXVII
FIGURE 143. START A NEW PROJECT IN CAMTASIA.................................................................................................... LXX
FIGURE 144. CAMTASIA RECORDING OPTIONS.......................................................................................................... LXXI
FIGURE 145. CAMTASIA RECORDING OPTIONS.......................................................................................................... LXXI
FIGURE 146. CAMTASIA RECORDING MODE.............................................................................................................LXXII




           AWPR – Automatic Wireless Phone Registration
                                                                                                                                   December 5, 2005


                                                                  List of Tables
TABLE 1. TECHNICAL MERIT. ........................................................................................................................................ 3
TABLE 2. SPI AND CPI CALCULATIONS........................................................................................................................ 21
TABLE 3. TEAM MEMBERS' CONTRIBUTION. ................................................................................................................ 23
TABLE 4. CALCULATED ISOTROPIC RADIUS VALUES.................................................................................................... 71
TABLE 5 . CALCULATED AND SOFTWARE-PREDICTION RADII AT VARIOUS RECEIVED SIGNAL LEVELS. ...................... 75
TABLE 6. CALCULATED CARDIOID RANGE VALUES. .................................................................................................... 77
TABLE 7. CALCULATED AND SOFTWARE-PREDICTED RANGES AT VARIOUS RECEIVED SIGNAL LEVELS. .................... 80
TABLE 8. STAR VISION’S MEASURED VS. PREDICTED RSSI VALUES. ........................................................................... 83
TABLE 9. ADVANTGX'S MEASURED VS. PREDICTED RSSI VALUES.............................................................................. 83
TABLE 10. WAVEGUIDE DETAILED SPECIFICATIONS. ................................................................................................... 86
TABLE 11. DETAILED ANTENNA SPECIFICATIONS. ....................................................................................................... 89
TABLE 12. VOIP TO GSM AUTOMATIC REGISTRATION ................................................................................................ 93
TABLE 13. GSM TO VOIP AUTOMATIC REGISTRATION TEST RESULTS........................................................................ 93
TABLE 14. SIGNAL STRENGTH TEST RESULTS.............................................................................................................. 94
TABLE 15. AVERAGE TEST RESULTS. ........................................................................................................................... 97
TABLE 16. DUTY CYCLES............................................................................................................................................. 98
TABLE 17. VOIP TO GSM AUTOMATIC REGISTRATION TEST RESULTS........................................................................ VII
TABLE 18. GSM TO VOIP AUTOMATIC REGISTRATION TEST RESULTS....................................................................... VIII
TABLE 19. VOIP TO GSM DEDICATED MODE SIGNAL STRENGTH TEST RESULTS........................................................ XII
TABLE 20. VOIP TO GSM IDLE MODE SIGNAL STRENGTH TEST RESULTS................................................................... XII
TABLE 21. GSM TO VOIP IDLE MODE SIGNAL STRENGTH TEST RESULTS................................................................... XII
TABLE 22. GENERATED TRAFFIC LEVELS ON IXIA TRAFFIC GENERATOR. .......................................................... XXXVIII
TABLE 23. VOICE QUALITY VS. GENERATED TRAFFIC. ................................................................................................ XL
TABLE 24. GSM IDLE MODE TIMES. ...................................................................................................................... LXXIV
TABLE 25. GSM DEDICATED MODE TIMES. ............................................................................................................ LXXV
TABLE 26. GSM IDLE MODE (WITH 802.11 RADIO OFF). ........................................................................................ LXXVI
TABLE 27. GSM DEDICATED MODE (WITH 802.11 RADIO OFF).............................................................................. LXXVI
TABLE 28. VOIP IDLE MODE................................................................................................................................. LXXVII
TABLE 29. VOIP DEDICATED MODE..................................................................................................................... LXXVIII
TABLE 30. AVERAGE TEST RESULTS. ................................................................................................................... LXXVIII
TABLE 31. DUTY CYCLES..................................................................................................................................... LXXVIII




           AWPR – Automatic Wireless Phone Registration
                                                                               December 5, 2005



I. Project Overview
A. General


In the past few years, the communications industry has expanded tremendously in the
development of different ways to make phone calls. Currently, there are a variety of voice
networks that interface with the public switched telephone network (PSTN), including Voice
over Internet Protocol (VoIP) and cellular networks. These networks are built on different
principles and practices to ensure the delivery of reliable voice services to end customers. The
increased use of both VoIP and cellular services has led some cell phone manufacturers to create
dual-enabled handsets, which are able to operate on either network. Since most developed
countries throughout the world already contain the infrastructure for these services, these
handsets will prove to be very useful to the average businessman or woman in these countries.
On the next level, however, business users also have a desire to seamlessly switch between the
two networks, meaning that the user would like their phone to automatically register to the proper
network without having to make any changes to their phone settings. Unfortunately, the
technology does not currently exist to allow a call handoff to occur in mid-conversation.
However, the current handsets and network equipment are capable of performing an automatic
call registration, but extensive testing needs to be completed before companies decide whether or
not to implement this type of service. Companies wishing to offer this type of service are
concerned with the time it takes a phone to automatically register with the correct network, the
VoIP voice quality on these phones compared to the cellular voice quality, and how traffic
loading on the network affects the voice quality of a conversation. In addition, there are also
concerns regarding whether or not the received signal strength has a dramatic impact on the
conversation and at which received signal strength level the phone will decide to switch between
networks to a different network.




       AWPR – Automatic Wireless Phone Registration                                           1
                                                                                   December 5, 2005



B. Background

Currently, service providers throughout the world are in the process of moving towards
convergence in which all forms of communication will utilize a single system. Some service
providers have been pushing for the new unified systems, while the others are now realizing that
they will have to converge in order to remain in competition. A number of businesses have
already begun replacing their traditional voice systems with VoIP systems, and residential users
have begun replacing their landlines with VoIP. Nowadays, almost all users, business or
residential, also utilize cellular services in addition to their traditional phone or VoIP services.
By observing the transition of the world’s communications, cell phone manufacturers noticed
that there was a need to develop a single device, which would allow users to place a call on a
wireless VoIP, or cellular network. Although this is great for users that are planning on using
one network a majority of the time, users want their phone to have the ability to automatically
register with an 802.11 network if there is a strong enough signal available to make a call. This
is due to the fact that VoIP calls are much cheaper than using minutes on a cellular plan.
Through the completion of the AWPR project, Ringer Communications has implemented the
necessary setup to make an automatic phone registration possible between an 802.11 and GSM
network. In addition, Ringer Communications has performed all of the necessary testing to prove
whether or not this process is a desirable business solution.


C. Technical Challenge

The primary challenge of the AWPR project was to allow the dual-mode phone to register to the
appropriate network. By appropriate network, Ringer Communications means that the phone
should automatically register to a wireless VoIP network when the user enters the building and
the user is not currently in middle of a cellular phone session. The user should not be able to
connect to the VoIP network from any point ten feet outside of the building. When the user is
outside this range, the phone should automatically register with the GSM cellular network. If the
user is in mid-conversation on the wireless VoIP network, and then he or she roams outside of
the building, then the wireless VoIP call should be dropped and the phone should automatically



       AWPR – Automatic Wireless Phone Registration                                             2
                                                                                   December 5, 2005


register with the cellular service provider.


The assessment of technical merit for Ringer Communication’s project was calculated using
Table 1. After estimating the value of each category, a total of 1.1 was determined for this
project. The first factor is fulfilled by the project scope definition. The integration and testing of
the HP iPAQ, access points, security switch, and PoE switch satisfies the criteria for the second
technical merit factor. The requirements for the third technical merit is met by the automatic
registration timing test, the automatic registration signal level test, the traffic loading voice
quality test, the various signal strength voice quality test, and the antenna coverage and
containment test. The preliminary analysis of the antenna system using the Wireless Valley
"SitePlanner" software meets the requirements for the analysis technical merit factor. The
fabrication of a prototype network for testing satisfies the seventh technical merit factor. In order
to meet the criteria for the eighth technical merit factor, a website and a whitepaper has been
turned over to the project sponsor.


                                       Table 1. Technical Merit.
                            Technical Merit Factor                                  Applied
                                                                        Weight
                                                                                    Weight
          1    Contains a clearly described and completely
                                                                          0.1         0.1
               understood technical challenge
          2    Contains a requirement for system integration              0.2         0.2
          3    Contains a requirement for system testing                  0.2         0.2
          4    Contains a requirement for analysis                        0.2         0.2
          5    Contains hardware design, development and test             0.3          0
          6    Contains software design, development and test             0.3          0
          7    Contains a hardware fabrication requirement,
                                                                          0.2         0.2
               typically a prototype
          8    Contains a requirement for documentation other
                                                                          0.2         0.2
               than the project report
          9    Contains a requirement for intellectual property
                                                                          0.1          0
               protection
                                                           Total          1.8         1.1




       AWPR – Automatic Wireless Phone Registration                                             3
                                                                               December 5, 2005




D. Benefits

Although many companies know that automatic wireless phone registration is possible, this
technology has not been extensively tested or implemented before this project was completed.
Extensive testing has been completed for the dual-mode system implementation and the results
from these tests have been published in the form of a white paper. This white paper will be used
by companies wishing to implement such a service to help them determine whether the service is
something that they would like to pursue. In addition, as the head of the Telecommunications
Department for Texas A&M University, the team’s sponsor has an interest in the future
possibility of providing similar services on the Texas A&M campus to both students and faculty.
This project’s implementation and testing has provided the sponsor with the first steps needed to
pursue the implementation of this type of service.



E. Technical Report Structure

This document outlines the details of Ringer Communication’s AWPR project. It begins with an
overview of the project’s scope; including any objectives of the project, what was and was not be
done in the project, and any assumptions that were made while planning the project. After that,
the document talks about the project requirements and how each of the requirements were met.
Next, the report includes a Management section which contains the project milestones,
deliverables, and EVP (Earned Value Process). Then, the document contains a Technical
Networking section that includes a discussion of a basic SIP call flow, the protocols used in the
project, the registration process, and various call scenarios that were analyzed throughout the
project. The next section will contain the Equipment section which contains a discussion of
project network equipment, how to configure the equipment and a brief discussion of the testing
devices. After that, there is a Software section that contains a brief overview of the software used
for the project. The next section is provides a discussion of the software validation and antenna
selection that was made using Wireless Valley's "SitePlanner". Next, this document contains a
brief discussion of the test process and results. The document ends with a Conclusions section, a


       AWPR – Automatic Wireless Phone Registration                                         4
                                                                             December 5, 2005


Recommendations for Further Research section, and a Glossary section. The appendix of this
document will include the following: test plans, test results, and team member resumes.




      AWPR – Automatic Wireless Phone Registration                                        5
                                                                                December 5, 2005



II. Project Scope
The objective of this project was to design and integrate a wireless VoIP network and to test a
dual-mode phone’s ability to automatically register with the designed VoIP network or a Global
System for Mobile Communications (GSM). The wireless VoIP network will have the capability
of communicating with VeriSign via a VPN connection to complete a SIP registration. While in
the range of the 802.11 wireless network, the phone's voice traffic will be carried over the VoIP
network. However, when the user leaves the range of the access point and drops below a set
signal threshold, the dual-mode phone will automatically register with a Global System for
Mobile Communications (GSM) cellular carrier. While within the range of the 802.11 network,
the user will have the ability to roam seamlessly between wireless access points. This project
will not provide the user with the ability to roam between the 802.11 and GSM networks in mid-
conversation. Once the dual-mode phone’s 802.11 signal drops below a defined signal threshold,
the VoIP session will be ended and the phone will automatically register with the cellular
network.


The call registration from one network to the other should appear seamless to the user, meaning
that the user will not have to do anything on their device to switch from the 802.11 to the cellular
environment. The implementation of this project will involve the installation and configuration
of an access point, security switch, PoE switch, and a number of antennas. Testing, which is a
major focus of the project, included a timing analysis of automatic call registration process, the
VoIP voice quality on these phones compared to cellular voice quality, the effects of traffic
loading on voice quality, the effects of received signal strength on voice quality, and the signal
strength at which the phone decides when to switch to the other network.


In the process of planning this project, a number of assumptions were made. First, Ringer
Communications made the assumption that the Internet 2 Technology Evaluation Center (ITEC)
lab would be able to obtain all of the equipment in a timely manner as stated in previous
meetings. All of the equipment except for the VPN router did arrive on time as expected. In
addition, the project team assumed that the VPN connection between Texas A&M and VeriSign
would be established before the group began the implementation of the project. The VPN


       AWPR – Automatic Wireless Phone Registration                                          6
                                                                            December 5, 2005


connection was not established until the sixth week of the project implementation. This was due
to the late arrival of the VPN router and the coordination of communication between Texas
A&M’s CIS (Computer Information Services) department and VeriSign. Third, Ringer
Communications assumed that VeriSign would have completed their end of the network
implementation prior to the project implementation. VeriSign’s end of the network was properly
configured at the beginning of the project implementation phase.




      AWPR – Automatic Wireless Phone Registration                                      7
                                                                              December 5, 2005



III. Functional Requirements

In order to complete the project, a number of functional and performance requirements that were
created by the sponsor had to be met. For simplicity, these requirements have been related to the
four major technical merit factors that the group focused on for this project. These requirements
can be seen in the following sections.

A. System Integration


First, before anything else could be completed in the project, all of the local network components
had to be integrated to work together. This involved the integration of the access points, PoE
switch, the security switch, and the antennas. Then, once the local network was integrated, the
ITEC test network had to be integrated with VeriSign’s network in order to allow calls to be
registered through VeriSign. VeriSign is connected to the Cellular Service Provider
(FirstCellular) via the SS7 network and all three entities are connected to the PSTN. A block
diagram for the network can be seen in Figure 1.




                                  Figure 1. Network Block Diagram.




      AWPR – Automatic Wireless Phone Registration                                         8
                                                                                December 5, 2005


The first functional requirement that is contained within the integration portion of the project was
that we must have the capability to connect to a Wi-Fi or GSM network using a single device.
Performance requirements that lay under this requirement are that the device must operate on
802.11b and the device must use a FirstCellular SIM card. The HP iPAQ 6315 only operates on
802.11b and the FirstCellular SIM card was provided by VeriSign. The next functional
requirement for the integration portion of the project was that the access points were going to
need to be able to be powered at various locations throughout the building. A performance
requirement that falls under this functional requirement is that the access points must use 802.3af
(a POE standard) in order to power the cables over Ethernet. The Nortel 2230 Access Ports and
the BayStack 460 PoE switch both had the ability to utilize 802.3af.


The next functional requirement that is contained within the integration phase of the project is
that handoffs should be able to be made between access points in mid-conversation without
dropping the media stream. Two performance requirements fall under this function requirement.
First, in order to have a seamless transition between access points, a centralized WLAN control
will have to be used to control all access points throughout the building. To meet this
requirement the Nortel 2270 Security Switch was used along with the Nortel 2230 Access Ports
which were running the LWAPP (Light Weight Access Point Protocol) protocol. The second
performance requirement was that the handoff had to occur in less than 20 milliseconds in order
to avoid dropping the VoIP video stream. This requirement was specified by the Nortel
documentation. Although the AP to AP handoff time was not measured, the seamless handoff
does work properly and therefore must be less than 20 ms.


The next functional requirement in the integration portion of the project was that a current VoIP
standard must be used in order to register the VoIP calls with VeriSign. The two performance
requirements that were located under this functional requirement were that the VoIP standard
should be a SIP VoIP protocol and the G7.11 codec should be used. SIP and G7.11u were used
by all of the networking equipment for this project. The next major functional requirement for
the project was that the local portion of the project’s network should be integrated with the PSTN
(Public Switched Telephone Network) and with VeriSign. The two performance requirements
that fall under the functional requirement should connect to the PSTN via ITEC’s PSTN/IP


       AWPR – Automatic Wireless Phone Registration                                         9
                                                                              December 5, 2005


gateway and to VeriSign through a VPN connection. The VPN connection is established;
however, the project network is connected to the PSTN via VeriSign’s PSTN/IP Gateway. This
was due to some incompatibilities that were existent between A&M and VeriSign’s VPN routers.


The final network functional requirement for the networking integration was that all the devices
within the local network should communicate with each other. Under this requirement, there are
two performance requirements. First the access points must receive address from a DHCP
(Dynamic Host Control Protocol) server while the rest of the devices should have static
addresses.


Next, there are a large number of antenna system requirements that had to be met. The first
functional requirement is that users should have decent 802.11 coverage throughout the building.
The performance requirement for this function requirement is that the signal level inside the
building should be -75 dBm or greater in 95% of the coverage area. This practice is suggested in
the Nortel Access Port 2230 documentation because it allows the user to connect at 11 Mbps
which gives the user an approximate throughput of 5 Mbps. Although a single call only uses 75
kbps, the additional throughput is necessary to avoid problems caused by additional traffic loads.
A graph of the user throughput versus signal strength can be seen in Figure 2. After performing a
calculation based on the building floor plan one can see that the approximate square footage is
15,582. Therefore, 14803 square feet should have a signal strength of -75 or greater. After
implementing the wireless system it is estimated that approximately 14466 square feet was
covered with a signal strength of -75 dBm or greater. This is 93% of the building coverage area;
therefore, this requirement was not met. The remaining portion of the building had a received
signal strength level between -75 dBm and -85 dBm.




      AWPR – Automatic Wireless Phone Registration                                         10
                                                                                   December 5, 2005




                        Figure 2. Average client throughput vs. Signal Strength.


The next antenna system functional requirement that was given to Ringer Communications by
the sponsor was the signal should be contained within the building. The performance
requirements for this functional requirement were that the signal strength should be less than -89
dBm at a maximum of ten feet outside the building. This signal level was chosen due to the fact
that the iPAQ has a receiver sensitivity of -89 dBm for a 1 Mbps connection. The distance was
chosen because the project may be implemented in dorms in the future and the project sponsor
does not want the signal leaking into neighboring dorms. After implementing the wireless
system, it appears that the furthest distance at which the signal level is -89 dBm is approximately
ten feet. This requirement has therefore been met by Ringer Communications.


The next functional requirement for the antenna system was that there should be minimal overlap
between access points located on the same or neighboring channels. One performance
requirement for the antennas was that they could not operate on channels 1, 6, or 11 since these
channels are currently being used by other entities in the research park building. The project
team implemented a system which uses only channels 3 and 9. In order to avoid co-channel
interference by the two access points located on the same channels, the access points were spaced
on separate floors.




       AWPR – Automatic Wireless Phone Registration                                          11
                                                                                 December 5, 2005


The final antenna functional requirement was that an antenna must be chosen that is compatible
with the Nortel Access Port. The two performance requirements under this functional
requirement were that the impedance of the antenna should be 50 ohms and that the antenna must
be able to connect to a reverse female TNC connection, which is located on the back of the
Nortel Access Ports. The project team chose a Trendnet 6 dBi patch antenna which had a 50
ohm impedance and a reverse SMA male connector. The difference between the connectors was
fixed by using a reverse SMA female to reverse TNC male connector.



B. Analysis


The analysis section of the project is quite a bit smaller than the rest of the project. The only
functional requirement for the analysis phase of the project is that the antenna systems should be
modeled in a software package before the antennas would be purchased. The first performance
requirement for the antenna analysis was that the Wireless Valley "SitePlanner" software would
have to be used since it was the only package available. The next performance requirement was
that the software would have to be validated in order to help determine the accuracy of the results
produced by it. The software was validated by simulating an isotropic and directional antenna in
free space to ensure that the simulations met the values predicted by formulas. Then a validation
was made using the existing building access points. The final performance requirement is that
the group would have to model several antennas in order to aid the sponsor in making an antenna
selection.




       AWPR – Automatic Wireless Phone Registration                                           12
                                                                                December 5, 2005



C. Testing

The testing portion of the project composes a majority of the work completed. Therefore, there
were a large number of requirements that had to be met in this phase of the project. The first
functional testing requirement was that a timing analysis had to be performed on the automatic
registration process. There were two performance requirements that fell under this functional
requirement. First, the team had to record the time at which the phone de-registered from the
GSM network and registered with the 802.11 network. Next, the second requirement was the
team had to measure the time between when the phone was removed from the 802.11 network
and when it joined the GSM network.


The next functional requirement for the testing phase of the project was to perform saturation
testing on the wireless network. Two performance requirements that were part of this functional
requirement were that the IxExplorer software should be used along with the IXIA traffic
generator in order to flood the back of one of the access points with traffic. In addition, the team
was required to measure the voice quality at various generated traffic levels until an MOS score
of one was reached. This test was performed as required.


The third functional requirement for the testing phase of the project was that the team had to
record the signal levels at which the device was removed from one network and registered to
another. The first performance requirement for this test was that the team use a software tool on
the iPAQ to constantly monitor the signal level on the 802.11 network. This requirement was not
met due to the inaccuracy of the iPAQ software. Instead, the group used a handheld spectrum
analyzer called the “Grasshopper”. Next, the second performance requirement is that the
timestamps be matched between the signal monitoring software to an Ethereal capture to make a
correlation between the signal strength and the registration process. Rather than doing this, the
group walked away or towards the building and measured the signal strength seen by the
“Grasshopper” when the device indicated that it switched networks.




       AWPR – Automatic Wireless Phone Registration                                          13
                                                                               December 5, 2005


The next functional requirement was that the voice quality should be measured as a relation of
signal strength. For this test, the only performance requirements was that the voice quality
should be measured at received signal strengths between -40 dBm and -90 dBm in increments of
10 dB. This test was completed as required.


The main functional requirement that had to be completed for the project was AP to AP handoff
testing within the local VoIP network. In this test, the team had to ensure that the media stream
was not dropped when handing off traffic from one access point to another.


The final project testing requirement was to measure the battery life of the dual-mode handset.
The main performance requirements for the battery life test was that the battery life had to be
measured for the phone in idle mode or dedicated mode while the phone is registered on either
network. In addition, another performance requirement is that the phone should be tested without
the VeriSign client to how this changes the battery life.



D. Additional Documentation


The additional documentation phase of the project contains two main functional requirements.
The first main functional requirement was that the team had to complete a white paper for the
sponsor upon completion. This white paper had to contain a summary of the testing that was
completed along with a summary of the test results. This document should be directed towards
management-level readers. The second functional requirement for the additional documentation
portion was that the team had to design a web page for the sponsor which provided a summary of
the project along with links to project documentation. Both of these requirements have been
completed




       AWPR – Automatic Wireless Phone Registration                                         14
                                                                               December 5, 2005



IV. Management
A. Milestones


Ringer Communications periodically completed various milestones throughout the course of the
AWPR project. These critical events aided in showing progress throughout the project. The two
timelines in Figure 3 shows all of the project’s milestones as well as their estimated completion
dates.




                                       Figure 3. Milestones.




         AWPR – Automatic Wireless Phone Registration                                      15
                                                                                December 5, 2005


The first milestone Ringer Communications completed, on October 7, 2005, was the selection of
an antenna. This antenna was used to provide wireless access inside of the building. This was
necessary in order for the network testing to be completed.


The next milestone completed by Ringer Communications was on October 25, 2005. This
milestone was the completion of the voice quality testing, which involves testing the voice
quality on the VoIP network in both a traffic-free and saturated environment. The completed
tests, along with documentation including the results and analysis from these tests, comprise the
complete voice quality testing.


Ringer Communications’ third milestone was the completion of the automatic call registration
time testing, on November 12, 2005. In order for this to be completed, multiple successful
registration transfers had to be made and the time required to accomplish them was recorded.
This completed test includes the preliminary and post documentation, as well as the test itself.


The fourth milestone Ringer Communications completed, on November 18, 2005, was the
completion of the automatic call registration signal level testing. This test involved successfully
completing multiple registration transfers and recording the signal level that triggered them to
occur. It included the preliminary documentation, testing results and analysis, and post
documentation.


The fifth milestone Ringer Communications completed was the final white paper. This was due
on December 2, 2005. The white paper was requested by the team’s sponsor as a summary of the
project’s accomplishments and findings.


Ringer Communications’ final milestone was the final demonstration, on December 5, 2005.
This is the final presentation of the completed project to the sponsor, advisor, and course
instructor.




       AWPR – Automatic Wireless Phone Registration                                           16
                                                                                    December 5, 2005


B. Deliverables


Throughout the AWPR project, Ringer Communications completed multiple deliverables for the
sponsor. These deliverables included configuration documentation, antenna selection,
preliminary testing documentation, traffic free test results, network saturation test results,
automatic call registration time test results, automatic call registration signal level test results,
battery life test results, a final white paper for the sponsor, and a final technical report. These
deliverables can be seen in a timeline format. Figure 4 shows all of Ringer Communications’
deliverables for September and October, 2005.




                            Figure 4. Deliverables for September and October.




       AWPR – Automatic Wireless Phone Registration                                              17
                                                                                December 5, 2005




The first deliverable Ringer Communications presented to the sponsor was the configuration
documentation. This documentation was completed on October 1, 2005. It included all of the
steps necessary to completely configure the security switch, Bay Stack switch, PSTN/IP gateway,
and HP iPAQ. These steps involved the physical side, actually connecting the equipment to the
network, as well as the software side, all of the configuration files used with the equipment.


The second deliverable presented by Ringer Communications was the antenna selection, on
October 7, 2005. This antenna was used to provide a wireless signal to the phone to allow VoIP
calls. It covered only inside the ITEC facilities.


The next deliverables, on October 10, 2005, was the preliminary test documentation. This
documentation verified that the network equipment was correctly configured.


The fourth deliverable, the traffic-free test results, was completed on October 19, 2005. In this
deliverable, Ringer Communications compiled all of the documentation about the traffic-free
voice quality test. This test essentially involved testing the quality of a VoIP call on the network
when there was no other traffic present. The documentation included the test plan that was
completed and the results obtained from the test. This documentation, similar to all of the other
test documentation, included graphs, tables, a full description of how to perform the test, and a
description on how to interpret the test results.


The fifth deliverable Ringer Communications presented, on October 25, 2005, was the network
saturation test results. This test involved testing the quality of a VoIP call on the network when
the access point was saturated with various types of traffic. The documentation included the test
plan that was completed and the results obtained from the test. Figure 5 shows the deliverables
for November and December.




       AWPR – Automatic Wireless Phone Registration                                          18
                                                                                 December 5, 2005




                          Figure 5. Deliverables for November and December.

The sixth deliverable Ringer Communications presented to the sponsor, on November 12, 2005,
was the automatic call registration time test results. This test involved making several successful
registration transfers from the VoIP network to the cellular network and back. The time required
to complete the transfer was recorded. The documentation for this deliverable included the test
plan that was completed and the results obtained from the test.


The seventh deliverable, completed on November 18, 2005, was the automatic call registration
signal level test documentation. In this deliverable, Ringer Communications compiled all the
documentation about the registration transfer signal level test, including the test plan that was
utilized and the results from the test. This involved completing several successful registration
transfers from the VoIP to cellular network and back. For each of these, the received signal level
that triggered the transfer was recorded and analyzed.


The eighth deliverable presented by Ringer Communications was the battery life test
documentation, on November 30, 2005. This test involved analyzing and recording the length of
talk time that the phone will be able to remain powered by its battery on both the VoIP and
cellular networks. These recorded times were then compared and analyzed. The documentation
for this deliverable included the test plan that was completed and the results obtained from the
test.




        AWPR – Automatic Wireless Phone Registration                                         19
                                                                               December 5, 2005




The ninth deliverable Ringer Communications presented was the final white paper for the
sponsor. This document was scheduled to be delivered to the sponsor on December 2, 2005. It
was be a summary of the work completed in the AWPR project. This included the various
discoveries made, as well as the overall view of the project. This document was submitted to the
sponsor as a bound hard copy with an included digital copy on a CD.


The last deliverable presented by Ringer Communications was the final technical report. This
document was scheduled to be completed on December 5, 2005. It was a compilation of all of
the work done in the entire AWPR project. It was to be presented as a spiral hard copy as well as
in a digital format on a CD included with the bound report.


C. Earned Value Process

The EVP (Earn Value Process) was a tool used by Ringer Communications to track the status of
the project from week to week. The EVP values were calculated weekly and were based on the
BWCP (Budgeted Costs of Work Performed), the BCWS (Budgeted Costs of Work Scheduled),
and the ACWP (Actual Costs of Work Performed). The calculations led to the CPI (Cost
Performance Index) and SPI (Schedule Performance Index). The SPI value allowed the TAT
team to quickly see if the team was on schedule. If this index value was below 1, then the project
team was behind schedule that week. Otherwise, if the SPI was above 1, then the project team
was ahead of schedule that week. The other index value, CPI, allowed the TAT team to
determine whether or not the project team was on budget. If the CPI value was less than 1, then
the project team was over budget; otherwise, if the CPI value was greater than 1, then the project
team was under budget. All of the EVP calculations for each week can be seen in Table 2 below.




      AWPR – Automatic Wireless Phone Registration                                         20
                                                                               December 5, 2005




                             Table 2. SPI and CPI Calculations.
                                                              Budgeted Cost of
        Budgeted Cost of Work    Actual Cost of Work
Week                                                              Work Schedule      CPI    SPI
          Performed (BCWP)        Performed (ACWP)
                                                                    (BCWS)
 1            138989.69                139385.625                   139547.5         1.00 1.00
 2            141325.94                 141901.25                    143275          1.00 0.99
 3            143727.81                144416.875                   146136.25        1.00 0.98
 4            145388.13                 151776.25                    148945          0.96 0.98
 5             146589.1                 154611.3                    150913.8         0.95 0.97
 6             151241.9                 157459.4                    152777.5         0.96 0.99
 7             154398.4                 160281.3                     157345          0.96 0.98
 8             155553.4                 163076.9                    160521.3         0.95 0.97
 9             158185                   166292.5                     163015          0.95 0.97
 10            158185                   166292.5                     163015          0.95 0.97
 11           160330.94                 168917.5                     166795          0.95 0.96
 12            162903.4                 171299.7                    168107.5         0.95 0.97
 13            164839.4                  170785                     176050.9         0.97 0.94
 14            174512.5                  172255                     178833.4         0.98 1.01




      AWPR – Automatic Wireless Phone Registration                                         21
                                                                                December 5, 2005


A graphical view of the SPI and CPI values can be seen in Figure 6.




                                     Figure 6. EVP (SPI vs. CPI).


As seen in the table and in the figure, the project team started off on schedule and on budget on
week 1 and remained slightly behind schedule and over budget for the remainder of the project.
Additional equipment had to be purchased for the project which caused the project to be over
budget. In addition, one of the tests that were scheduled to be performed, dropped packets
testing, was cancelled early in the project due to its irrelevance. This caused the project to appear
behind schedule. Upon completion, the project had a SPI of approximately one.




D. Individual Contributions


Although some of the work that was completed throughout the duration of the project involved
all four team members simultaneously, each team member was responsible for leading specific
tasks. The project manager, John Kelley, coordinated efforts between VeriSign, the ITEC staff,


       AWPR – Automatic Wireless Phone Registration                                          22
                                                                                 December 5, 2005


and the project team. This involved arranging meetings, the ordering of equipment, and
arranging the scheduling of tests. Mr. Kelley also led a majority of the documentation efforts
throughout the project including the authoring and editing of almost all of the team’s documents.
In addition, Mr. Kelley assisted the project team with various other tasks as needed. Next, the
quality test engineer, Chris Rue, led the voice quality testing tasks. Within the voice quality
testing there were three separate major tests including: signal strength, wired traffic loading, and
wireless traffic loading. Mr. Rue also led the automatic registration timing test and the
installation of the antennas throughout the building.


Kara Flanary, one of the network engineers, led the network configuration phase of the project.
This involved setting up and integrating the BayStack switch, the Nortel Security Switch, the
access points, and the iPAQ. In addition, Ms. Flanary also led the battery life testing efforts.
The other network engineer, Ranesha Cobb, led the antenna analysis phase of the project. In
order to do this task, Ms. Cobb had to learn the Wireless Valley "SitePlanner" software package.
In addition to this contribution, Ms. Cobb also designed and maintained the team’s website and
ITEC’s webpage concerning the AWPR Project.


Table 3 shows the total number of planned hours and the total number of hours worked for each
team member.


                                Table 3. Team Members' Contribution.
               Position              Team            Total Planned         Total Hours
                                    Member               Hours               Worked
          Project Manager        John Kelley               379                464.75
          Quality Engineer       Chris Rue                 341                 353.5
          Network Engineer       Ranesha Cobb              361                 305.5
          Network Engineer       Kara Flanary              367                 320.5
                      Total Hours                         1448                1444.25




       AWPR – Automatic Wireless Phone Registration                                          23
                                                                                 December 5, 2005


V. Technical Networking

The overall network that was used in order to implement this project is very complex and
involves three major entities. A detailed view of the network can be seen in Figure 7.




                                     Figure 7. Functional Design.


Although this network may seem somewhat confusing, the following sections will help explain
how the network operates. First, the document will begin with a discussion of a basic SIP call
flow. Then, there will be a discussion of the various protocols used to allow communication in
the project. After that, there will be a discussion of the call registration process that the iPAQ
must undergo every time the phone is registered on the VoIP network. The Technical
Networking section will end with a detailed discussion of call flows may occur in the existing
network.




       AWPR – Automatic Wireless Phone Registration                                           24
                                                                                December 5, 2005


A. Basic SIP Call Flow


Before the network can be understood, one must gain a basic understanding of the SIP protocol
and how the call setup process works. Initially, before a phone can operate on a network, it must
register to a SIP server. Registration must be completed so that the server knows the device is on
the network. The phone will send a registration request to the server and the server should send
an OK message back to let the phone know that it successfully registered. Once the phone is
registered, it can make or receive calls on the VoIP network. An example call setup can be seen
in Figure 8.




                                 Figure 8. SIP Call Setup\Breakdown.


First, the VoIP phone must send an INVITE message to the remote party’s SIP server. Once, the
remote SIP server receives the INVITE, it will send the INVITE message to the remote party and
in the meantime it will send back a 100 status message back to the calling party to indicate that is
trying to connect the user to the remote party. Once the remote party picks up the phone, the
remote SIP server will send an OK message back to the calling party to let it know that the call is



       AWPR – Automatic Wireless Phone Registration                                         25
                                                                                 December 5, 2005


connected. Then, an RTP stream will travel in between the two parties as long as the call is in
progress. The RTP packets contain the audio information for each of the parties. After the users
have completed their call, one of the parties will hang up first. After the user hangs up, the phone
will send a BYE message to the SIP server and the SIP server will send the BYE message to the
appropriate party. The party that receives the BYE message will send an OK back to the SIP
server to let the party know that the call has been terminated.

B. Project Protocols


In order for all of the devices to communicate properly with one another, a number of protocols
had to be utilized in the implementation phase of this project. Each of these protocols
corresponds with one or more layers of the OSI (Open System Interconnection) model. The
wireless VoIP protocol stack that was used in this project can be seen in relation to the OSI
model in Figure 9.




                       Figure 9. OSI Model Compared to Project Protocol Stack.




       AWPR – Automatic Wireless Phone Registration                                         26
                                                                                 December 5, 2005


As seen in the figure the G.711u VoIP codec was located at the top of the wireless VoIP protocol
stack. The codec relates to the application and presentation layers of the OSI model. The
application layer of the OSI model supports application and end-user processors. The application
layer of the OSI model handles the application and end-user processes that are necessary for the
management of a device on a network. Specifically, this layer may deal with the authentication
and privacy of a device on a network. The other layer in which the voice codec fall into is layer
six, the presentation layer. This layer handles the translating which must take place between an
application and a network. Encryption can also take place at this layer of the network.


Next, the protocols SIP and SDP (Session Description Protocol) that are used in the project fall in
the fifth layer of the OSI model, the session layer. SIP is the protocol which is used to register a
device on a VoIP network and it also handles the signaling necessary for call session setup and
termination. The other protocol in this layer, SDP, aids the SIP protocol by sending specific
session information to nodes which are being connection. This protocol can be used to transfer
the following information to a device regarding the connection which is being established:
timing, format, the name of the session, the codec, as well as a number of other things.


Next, the project contains three separate protocols which operate in the fourth layer of the OSI
model, the transport layer. First, the RTP protocol is used to transport the actual audio from a
VoIP call. This protocol typically runs on top of UDP, a connection-less transport protocol in
order to utilize the multiplexing and checksum abilities of UDP. The third protocol, RTCP (RTP
Control Protocol), is used to monitor the delivery of the RTP packets and allows a receiver to
make adjustments for packet loss and other network impairments. By using all three of these
protocols, the VoIP network is able to provide end-to-end delivery of data and provide flow
control throughout the network. This is the basic definition of the transport layer of the OSI
model and that is why all three protocols fall within this layer.


As one can clearly see the IP protocol and the LWAPP protocol are two protocols which fall
within layer 3 of the OSI model, the network layer. The IP protocol handles the routing and
addressing of the network components. The LWAPP (Light Weight Access Point Protocol) is
used by the access points to communicate with the Nortel switch. This is a fairly new protocol,


       AWPR – Automatic Wireless Phone Registration                                          27
                                                                                 December 5, 2005


but it allows a wireless network to be centrally managed. When using this protocol, an access
point will automatically attempt to find and connect to a security switch when it is plugged in.
Once the access point finds the security switch, it uses the protocol to download its settings and
any updates that need to be made.


Layer 2, the data link layer is split into two portions, the MAC (Media Access Control) and LLC
(Logical Link Control). The LLC layer handles frame synchronization, flow control, and error
checking. For the LLC portion of layer 2, the wireless VoIP network uses 802.2, the Ethernet
standard. The 802.11 wireless standard which is used in the AWPR project utilizes the MAC
portion of the data link layer and the physical layer. The MAC sub-layer is used to controls how
a network gains access to the network and the physical layer is used to translate the electrical
signal into a bit stream.


C. Call Registration

Before one of the VoIP devices communicates on the network it must first register itself on a
server. For the purpose of this project the wireless phones had to register to a SIP server which
was located at VeriSign’s location. This server, the NRD, had to be used because it had access to
VeriSign’s Wireless Mobile Gateway which can communicate with a cellular service provider.
The connection to the cellular service provider is essential for this project since the device will
need to switch between GSM and VoIP. In addition to device authentication, the NRD is used to
track a device’s current location.


In order to begin the registration process, the phone will automatically send out an SIP
registration message. The interval at which the phone sends out these SIP registration messages
can be set by the user and can be set as low as one second. If the phone is configured correctly,
the phone will send the SIP registration request message to VeriSign’s SIP server. Upon receipt
of this message, VeriSign’s SIP server will send a message to the Wireless Mobile Gateway in
order to specify that the user has requested to register to the VoIP network. When the Wireless
Mobile Gateway receives this information it will send a message to the cellular network provider,
FirstCellular, to inform the provider that the user has moved off of their network and onto the


       AWPR – Automatic Wireless Phone Registration                                           28
                                                                                December 5, 2005


VoIP network. FirstCellular will then change its HLR to indicate that the user is roaming on
VeriSign’s network. After changing this entry, the cellular service provider will send a message
to VeriSign’s Wireless Mobile Gateway to confirm that the entry has been changed. When the
Wireless Mobile Gateway gets this message it will send a confirmation to the SIP server to let the
server know that the entry has been changed. Next, the SIP server will then send an OK message
back to VoIP device which sent the registration request to let it know that is has successfully
registered to the VoIP network. Upon receipt of the OK message from the SIP server, the phone
will switch to VoIP mode and will be able to send and receive message on the VoIP network
using the same phone number.


D. Call Scenarios

As one can imagine, receiving and making a call using the phone is much more complicated than
the registration process. In order to help explain the call flow which takes place for different
types of calls, the project team has created a number of call scenarios which will be discussed in
greater detail.


The first and easiest scenario to understand is when a call is made from a phone on the PSTN to
one of the dual-mode handsets when it is on the cellular network. This call flow scenario can be
seen in Figure 10.




       AWPR – Automatic Wireless Phone Registration                                          29
                                                                            December 5, 2005




                                 Figure 10. PSTN to GSM Call Scenario.


The call flow occurs in the following order:
   1. The PSTN phone originates the call and it travels to the PSTN.
   2. The call then travels from the PSTN through the SS7 signal network to the cellular
       network provider.
   3. The cellular network provider then sends a message to the dual-mode phone to indicate
       that a party is trying to reach it.
   4. When the phone picks up the phone conversation, a media stream will be established
       between the dual-mode phone and the PSTN phone.


   Next, a more complicated scenario is when one of the dual-mode phone’s in VoIP mode
   makes a call to the other dual-mode while it is in GSM mode. A picture of this scenario can
   be seen in Figure 11.




      AWPR – Automatic Wireless Phone Registration                                         30
                                                                             December 5, 2005




                                Figure 11. VoIP to GSM Call Scenario.


The VoIP to GSM call flow occurs in the following order:
   1. The dual-mode VoIP phone sends an INVITE to VeriSign’s SIP server.
   2. The call leaves VeriSign’s network via the PSTN/IP gateway and it travels to the PSTN.
   3. Next, the PSTN sends the signaling to the cellular network provider.
   4. The cellular network provider then send a message to the dual-mode phone in GSM mode
       to indicate that someone is trying to contact it.
   5. The user on the GSM dual-mode phone answers the phone and a media stream is
       established between the VoIP phone and the GSM phone.


   Next, the VoIP to VoIP scenario is a little more complex because it involves signaling
   between VeriSign and the cellular company. A picture of a VoIP to VoIP call can be seen in
   Figure 12.




      AWPR – Automatic Wireless Phone Registration                                          31
                                                                               December 5, 2005




                               Figure 12. VoIP to VoIP Call Scenario.


The VoIP to VoIP call flow occurs in the following order:
   1. The dual-mode VoIP phone sends an INVITE to VeriSign’s SIP server.
   2. VeriSign’s PSTN/IP gateway sends the call to the PSTN.
   3. The call will then travel from the PSTN to the cellular service provider.
   4. Upon arrival at the cellular service provider, the provider will look up the phone in its
       HLR and realize that the phone is not on its network. The cellular service provider will
       send a request to VeriSign’s Wireless Mobile Gateway in order to retrieve information on
       how the call can be routed to its current location. When the Mobile Gateway receives this
       receives this request it will communicate with the SIP server to open up a window to
       allow an incoming call to be routed to the phone.
   5. Next, the Mobile Gateway will send the cellular service provider back the routing
       information.
   6. The cellular service provider will then route the call back to PSTN towards VeriSign’s
       PSTN/IP gateway.
   7. The call will travel from the PSTN to VeriSign’s gateway.




      AWPR – Automatic Wireless Phone Registration                                         32
                                                                               December 5, 2005


   8. The PSTN/IP gateway will route the call to the called dual-mode handset which is in
       VoIP mode.
   9. When the phone picks up the call, a media stream will be established between the called
       VoIP phone and the calling VoIP phone.


   The final call scenario, PSTN to VoIP, is very similar to the VoIP to VoIP call scenario. It
   can be found in Figure 13.




                                Figure 13. PSTN to VoIP Call Scenario.


The PSTN to VoIP call flow occurs in the following order:
   1. The PSTN phone sends a call to the PSTN.
   2. The call will then travel from the PSTN to the cellular service provider.
   3. Upon arrival at the cellular service provider, the provider will look up the phone in its
       HLR and realize that the phone is not on its network. The cellular service provider will
       send a request to VeriSign’s Wireless Mobile Gateway in order to retrieve information on
       how the call can be routed to its current location. When the Mobile Gateway receives this
       receives this request it will communicate with the SIP server to open up a window to
       allow an incoming call to be routed to the phone.



      AWPR – Automatic Wireless Phone Registration                                         33
                                                                         December 5, 2005


4. Next, the Mobile Gateway will send the cellular service provider back the routing
   information.
5. The cellular service provider will then route the call back to PSTN towards VeriSign’s
   PSTN/IP gateway.
6. The call will travel from the PSTN to VeriSign’s gateway.
7. The PSTN/IP gateway will route the call to the called dual-mode handset which is in
   VoIP mode.
8. When the phone picks up the call, a media stream will be established between the called
   VoIP phone and the calling VoIP phone.




   AWPR – Automatic Wireless Phone Registration                                        34
                                                                               December 5, 2005



VI. Equipment
A. Network Equipment
For the AWPR project to function properly, several pieces of network equipment were necessary.
This included a Nortel 2230 Access port, a Nortel BayStack 460 Switch, a Nortel WLAN
Security Switch 2270, and an HP iPAQ 6315.




                                 Figure 14. Equipment at ITEC Lab.


1. Nortel 2230 Access Port


The Nortel 2230 Access Port is used to allow the iPAQ phones to connect wirelessly to the
802.11 network. This is the device that the antenna connected to. It was responsible for
propagating the signal and allowing wireless devices to access the wired network. Three access
ports were used in the final network implementation. The placement of these devices coincides
with the placement of the three antennas. Due to the fact that each of these devices is a dumb
device, they must be controlled by a central security switch. This switch will be discussed in part
3.




      AWPR – Automatic Wireless Phone Registration                                         35
                                                                                December 5, 2005


2. Nortel BayStack 460 Switch


The Nortel BayStack 460 Switch is used to connect the test network to the ITEC network and to
provide Power over Ethernet (PoE) to the Nortel Access Ports. The BayStack switch was
physically connected to the security switch, the access ports, and the ITEC Internet router. This
is how it connected the AWPR network to the ITEC network. Since the access ports do not have
power available to them in the ceiling where they are installed, this power must be supplied
through other means. The method the Nortel Access Ports utilize is Power over Ethernet. This
means that power is sent to the access ports over the unused pairs in the category 5e cables.



3. Nortel WLAN Security Switch 2270


The Nortel WLAN Security Switch 2270 is used to control the Nortel Access Ports. This is the
device that controls the various options pertaining to the wireless signal. One such option is the
various security features. The security switch allows for various types of security to be
implemented over the wireless network. The one chosen for the AWPR project was WEP. This
device can also be used to control the power usage of the access ports. For the purposes of this
project, the power was set to the lowest setting. This was to limit how far the signal propagated.
Finally, the security switch can be used to control wireless provisioning.


4. HP iPAQ 6315


The HP iPAQ 6315 is the dual-mode phone used to connect to the wireless network. This phone
was chosen because it has three different wireless radios. These radios include Bluetooth, GSM,
and 802.11. The GSM radio allows the phone to connect to a cellular network in order to make
normal cellular calls. The 802.11 radio allows the iPAQ to connect to the wireless 802.11g. A
soft client, installed on the iPAQ, allows it to make Voice over IP calls over the wireless
network. This dual-mode capability made the phone ideal for the AWPR project.




       AWPR – Automatic Wireless Phone Registration                                           36
                                                                              December 5, 2005


B. Equipment Configurations


1. Nortel BayStack 460 Switch


The Nortel BayStack 460 Switch can be configured by using the web interface. In order to
access this interface, a computer must be connected directly to the switch. This computer must
be logically located on the same subnet as the switch. The computer will be located on the same
subnet if its IP address is on the same subnet. The switch’s IP address is 165.91.82.99 with a
mask of 255.255.255.128. Any IP less than .128 will be located on the same subnet. The web
interface can then be reached by entering a web browser and typing http://165.91.82.99 into the
explorer bar. The web interface can be seen in Figure 15.




                               Figure 15. BayStack 460 Web Interface.


As the first step of configuring the BayStack, the PoE settings should be completed. This can be
done by entering the Power Management option under the Configuration menu. First, the Global
Power Management settings should be set. This screen can be seen in Figure 16.




      AWPR – Automatic Wireless Phone Registration                                         37
                                                                                December 5, 2005




                               Figure 16. PoE Global Power Management.


The first three fields are indication fields which allow the user to know the status of the PoE
functions. The Available DTE Power field displays the amount of power that is available for
powering devices over Ethernet. The DTE Power Status field displays whether or not all of the
Power over Ethernet functions are working correctly across the entire switch. Normal indicates
that everything is functioning properly whereas Error-PoE failed indicates an error. The DTE
Power Consumption field indicates how much power is actually being drawn from the security
switch.


Next, the user can specify some of the global power settings. First, the user can specify the DTE
Power Usage Threshold. If the devices connected to the PoE switch exceed this threshold, then
the switch will trigger a trap command. For this project, the default value of 80% has been
chosen. Next, the Power Pair option allows the user to specify which pins of the RJ-45 should be
used to supply the power. If the user chooses spare the PoE switch will use the pins which are
not being used for data transmission. If the user chooses the signal option then the PoE switch
will use pins which are also being used for data transmission. Since the network used in this
project is only 100Mbps, there are four spare pins to be used for PoE; therefore, the Power Pair
option will be set to spare.




       AWPR – Automatic Wireless Phone Registration                                          38
                                                                                 December 5, 2005


After that, the user can enable or disable traps from the agent. This determines whether or not a
system interrupt can be called by a specific DTE device on one of the switch ports. This option
will be enabled for this project. The next option, PD Detect type, allows the user to specify what
type of devices will be powered by the switch. In order to make sure that the access points are
powered, this option has been set to 802.3af and legacy. The power standard used by Nortel’s
access points is 802.3af. The next fields in the Global Power Management configuration are the
DC Source Type and DC Source Configurations. The DC Source Type will be set to BayStack
10 even though that equipment will not be used for the project. There is not a need to have an
external power source for the purpose of this project, but the PoE switch still makes the user
specify what may be used. The DC Source Configuration will be left at the default Power
Sharing option even though there is not an external power source.


Next, the user can configure specific PoE settings for each port. These options can be seen in the
Port Property screen under the Power Management sub-menu. This can be seen in Figure 17
below.




                               Figure 17. PoE Port Power Management.


First, the user can specify whether or not PoE is enabled on each port. Next, the user can see the
status of the PoE function for each port. If nothing is connected to the port, then the status field
will say Detecting. If a non-PoE device is plugged into a port, it will say Invalid PD and if a
valid 802.af device is plugged into the port, the status field will say delivering power. In this


         AWPR – Automatic Wireless Phone Registration                                         39
                                                                               December 5, 2005


configuration the user can also configure a maximum power limitation, the priority of the PoE
feature compared to other devices. The final three fields on this configuration display the current
voltage, current, and power being used on that particular port.


After completing the PoE configuration the user should setup the VLANs on the switch. This
can be done by clicking the Application option on the main menu and then selecting the VLAN
Configuration option under the VLAN sub-menu. After this is done, the screen shown in Figure
18 will be displayed.




                            Figure 18. BayStack 460 VLAN Configuration.


In the VLAN configuration screen, the VLAN table lists the current VLANs. The X button will
allow a user to delete the VLAN while the list button will allow a user to edit the current VLAN.
The VLAN Creation field will allow a user to create a VLAN with specific assigned ports. This
specific process will be further defined in the next few steps. The VLAN Setting option will
allow the administrator to set which VLAN will be the management VLAN. This is the VLAN
that will be allowed to access to the web based interface. The final VLAN Configuration option




      AWPR – Automatic Wireless Phone Registration                                         40
                                                                              December 5, 2005


is the Auto PVID setting which will allow the Port VLAN Identifier, PVID, to automatically be
configured as a port is switched to a different VLAN.


To add a VLAN, should click on the Create VLAN button. Once this step is completed the user
should arrive at the screen shown in Figure 19.




                           Figure 19. BayStack Port-Based VLAN Creation.


The VLAN number on this screen is an identifier that can range from 1 to 16. The VLAN name
can be set to allow the administrator or user to easily identify the VLAN. The learning constraint
is set to IVL by default. Once the user is done setting the various VLAN settings, he or she
should click the Submit Button


The user will then be brought to a screen which will allow him or her to specify which ports
should be associated with the new VLAN. An example of the port association screen can be
found in Figure 20.




      AWPR – Automatic Wireless Phone Registration                                        41
                                                                             December 5, 2005




                            Figure 20. BayStack VLAN Port Association.


Once in this screen, the user can add the desired ports to the newly created VLAN. The user
should simply check the boxes corresponding to the desired ports. Then the user should click the
Submit button. To complete the VLAN configuration the user should on the Port Configuration
option located under the VLAN sub-menu. This screen is shown in Figure 21.




      AWPR – Automatic Wireless Phone Registration                                       42
                                                                              December 5, 2005




                           Figure 21. BayStack VLAN Port Configuration.
The Port Configuration screen allows the administrator to manually configure the PVID for each
port. This option is useful when the automatic PVID is turned off. If the user sets the PVID to 1
then the port will be set to work on VLAN 1. This concludes the configuration of the Nortel
BayStack 460 switch.


2. Nortel WLAN Security Switch 2270


In order to configure the WLAN Security, the web interface should be used. The user can
connect to the web interface of the Security Switch by typing in the address 165.91.82.100 in the
web interface. The user will should reach the screen shown in Figure 22.




      AWPR – Automatic Wireless Phone Registration                                        43
                                                                               December 5, 2005




                               Figure 22. Nortel Security Switch Login.


Once the user reaches the screen, he or she should login to the device and arrive at the main
screen. At the main screen, the user click on the WLAN option located on the main horizontal
menu located at the top of the screen. This will bring the user to the WLAN screen shown in
Figure 23




                       Figure 23. Security Switch WLAN Configuration Screen.


Once the user arrives at the WLAN configuration screen, he or she can click on the New…
Button located in the upper right hand corner of the screen. This will cause the screen shown in
Figure 24 to appear.




      AWPR – Automatic Wireless Phone Registration                                         44
                                                                       December 5, 2005




                       Figure 24. Security Switch New WLAN Screen.


This screen displays the basic settings that can be configured for the wireless network.
The WLAN should be named AWPR since this is the name of the project. The Radio
Policy option should be set to 802.11b/g only since 802.11b will be used for the test
network. The Layer 2 Security option should be set to Static WEP and the key can be
entered at the bottom of the screen. For the purpose of this project a 40 bit WEP key with
an index of 1 should be used. The WEP key should be abcdef1234. Next, the DHCP
Server option should be set to Override which force the access points on the AWPR to get
an address from a specific DHCP Server. The IP Address for the DHCP Server should be
set to 165.91.82.47.


Next, the Switch Ports for the Security Switch should be configured. In order to do this,
the user should click on the Switch option which is located on the main configuration
menu. Then, the user should click on the Ports option on the side bar and then he or she
should choose the Configure option. This screen can be seen in Figure 25.


AWPR – Automatic Wireless Phone Registration                                        45
                                                                          December 5, 2005




                        Figure 25. Security Switch Port Configuration.


The only change that should be completed in this portion of the configuration is the Physical
Mode of Port 1. This option should be set to 1000 Mbps Full Duplex in order to allow the
security switch to work with the BayStack 460 switch. Next, the individual access points that
are seen by the security switch should be configured. This can be done by selecting the
Wireless option from the main horizontal menu. This will bring the user to the screen shown
in Figure 26.




   AWPR – Automatic Wireless Phone Registration                                        46
                                                                                 December 5, 2005




                             Figure 26. Security Switch Wireless Settings.


The user should then select one of the access points and click on the Details option. This will
bring the user to the screen shown in Figure 27.




                        Figure 27. Security Switch Access Point Configuration.




      AWPR – Automatic Wireless Phone Registration                                         47
                                                                                 December 5, 2005


In this portion of the configuration, the user can specify the name of the access point. The name
should be somewhat descriptive of the access point’s location so that it will be easier to
determine which access point a client is connecting to. After the name has been changed for the
access point, the user should setup the RF settings for the wireless network. In order to do this,
the user should select the 802.11 b/g Radios option from the side menu. The user should then
select the desired access point from the list and click the Edit button. This will bring the user to
the screen shown in Figure 28.




                              Figure 28. Access Point Radio Configuration.




       AWPR – Automatic Wireless Phone Registration                                           48
                                                                              December 5, 2005


On the right side of the screen there is a section called RF Channel Assignment. Under this
heading, the user should select the Custom option for the Assignment Method. Then, the drop-
down menu should be used to select the desired channel. Finally, to finish the Security Switch
Configuration, the user should configure the supported and mandatory data rates. In order to this,
the user should select the 802.11 b/g Network option under the Global RF sub-menu. This will
cause the screen in Figure 29 to appear.




                           Figure 29. Security Switch Supported Data Rates.


The user should make sure that both the 802.11b/g Network Status box is checked. Then, a list
of available data rates should be displayed. The user should change 1 Mbps, 2 Mbps, 5.5 Mbps,
and 11 Mbps to mandatory. Although this would normally mean that the client would have to be
able to connect at those speeds, the iPAQs work differently than normal devices. If the desired




      AWPR – Automatic Wireless Phone Registration                                        49
                                                                                December 5, 2005


rates are not set to Mandatory, then the iPAQ is not even able to see the rates. After these rates
are setup, the user has completed the Nortel 2270 Security Switch configuration.


3. HP iPAQ 6315


As the first step of configuring the HP iPAQ for the project, the VeriSign PCTEL software
should be installed. In order to do this, the user should click on the Start Menu icon in the upper
left hand corner of the screen. Then the user should select the Programs option as shown in
Figure 30.




                                   Figure 30. HP iPAQ Start Menu.


       Once in the Programs window, the user should click on the File Explorer icon as shown
in Figure 31.




       AWPR – Automatic Wireless Phone Registration                                         50
                                                                                December 5, 2005




                             Figure 31. HP iPAQ Launching File Explorer.


Once in the File Explorer program, the user should click on the drop-down menu located next to
the PDA icon. In this menu, the user should select the My Device option. This will bring up a
listing of all of the main folders located in the root directory of the iPAQ. The user should
browse to the My Documents folder and then to the Personal folder. This will bring the user to
the screen shown in Figure 32.




       AWPR – Automatic Wireless Phone Registration                                         51
                                                                                       December 5, 2005




                    Figure 32. HP iPAQ Executing the VeriSign Soft Client Installer.


The user should then click on the PPC executable file to run the VeriSign installer. Once the
installation completed, the iPAQ will prompt the user to reboot the iPAQ. After a reboot has
been completed, the user will now be able to configure the Universal Phone (VeriSign’s PCTEL
soft client). The user can begin the configuration by launching the Start Menu and selecting the
Universal Phone option, as seen in Figure 33.




      AWPR – Automatic Wireless Phone Registration                                               52
                                                                              December 5, 2005




                         Figure 33. HP iPAQ Launching the Universal Phone.


Once in the Universal Phone application, the screen shown in Figure 34 should be displayed.
Note that the phone shows in the upper left hand corner that it is connected to the GSM network.
This will be useful in determining which network the phone is connected to.




                         Figure 34. HP iPAQ Universal Phone in GSM Mode.




      AWPR – Automatic Wireless Phone Registration                                       53
                                                                                 December 5, 2005


Next, the user should click on the WiFi icon located at the bottom of the screen. This will bring
the iPAQ to the screen seen in Figure 35.




                       Figure 35. HP iPAQ Universal Phone Wi-Fi Configuration.


Once in the WiFi setup screen, the user should select the AWPR network. Next, the user should
click the Edit Profile icon located at the bottom of the screen. This icon is highlighted in the
previous figure. After the user clicks on the icon, the screen shown in Figure 36 will be
displayed.




       AWPR – Automatic Wireless Phone Registration                                          54
                                                                                 December 5, 2005




                    Figure 36. HP iPAQ AWPR Wireless Network General Settings.




On this screen, the user should change the Connection options to Automatic and then the user
should click on the Security Tab located at the bottom of the screen. This will cause the screen
shown in Figure 37 to appear.




      AWPR – Automatic Wireless Phone Registration                                         55
                                                                                  December 5, 2005




                    Figure 37. HP iPAQ AWPR Wireless Network Security Settings.


In the Security screen, the user should click on the box to enable data encryption. Then the user
should enter the AWPR network key into the WEP Key field. Once this is complete, the user
should click the OK button located at the upper right hand corner of the screen. This will bring
the user back to the WiFi screen as shown in Figure 38.




                    Figure 38. HP iPAQ Connecting to an AWPR Wireless Network.



      AWPR – Automatic Wireless Phone Registration                                          56
                                                                               December 5, 2005




Next, to connect to AWPR the user should click the Green play button. Once the iPAQ has
successfully connected to the network, the user should see the screen shown in Figure 39.




                Figure 39. HP iPAQ Successful Connection to AWPR Wireless Network.


Now the user should return to the main Universal Phone screen. This can be done by clicking on
the phone icon located at the bottom of the window. The phone icon is depressed in the previous
picture. After the user clicks on the icon, the screen shown in Figure 40 will be displayed.




      AWPR – Automatic Wireless Phone Registration                                          57
                                                                                  December 5, 2005




          Figure 40. HP iPAQ Universal Phone Attempting to Connect to VeriSign's SIP Server.


Notice that the application says VoIP in the upper-left hand corner with three dots below it. This
is indicating that the phone is trying to connect to the VoIP network. After a couple of seconds,
the phone’s registration will fail and the user should see the screen shown in Figure 41.




                       Figure 41. HP iPAQ Universal Phone SIP Error Message.




      AWPR – Automatic Wireless Phone Registration                                             58
                                                                               December 5, 2005




The “SIP server unreachable” error message indicates that the phone cannot communicate with
VeriSign’s SIP server. This error message should be received upon the initial attempt to connect
to VeriSign’s server because the network parameters for a SIP call have not been setup yet.
In order to setup the VoIP settings, the user should click on the Tools Menu icon at the bottom of
the screen and then he or she should select the Settings option followed by the Configuration
option. This process can be seen in Figure 42.




                         Figure 42. HP iPAQ Starting the SIP Configuration.


By performing these actions, the user will arrive at the screen shown in Figure 43.




      AWPR – Automatic Wireless Phone Registration                                         59
                                                                                December 5, 2005




                        Figure 43. HP iPAQ Universal Phone SIP Configuration.


Within this setup screen, the user should enter 209.112.113.19 in the Proxy Server and Domain
fields. This is the IP address of the VeriSign SIP server. Next, the user should enter
31901000000123X in the username field. The X will be different depending on which phone the
user is using. One phone will have a 4 in place of the X and the other phone must use a 3. The
Register Timeout field can be set to whatever the user would like. This is how often the phone
will attempt to register with VeriSign’s SIP server. By default, this value is set to 120, meaning
that the phone will attempt to register every 120 seconds. For the purpose of this project, this
value is changed to the minimum time of 1 second. The Audio options should be left at the
default settings. The completed setup for one of the phones used in this project can be seen in the
previous figure.


Next, the user should click on the Authentication option in the setup screen. This will bring the
user to the screen shown in Figure 44.




       AWPR – Automatic Wireless Phone Registration                                         60
                                                                                  December 5, 2005




                    Figure 44. HP iPAQ Universal Phone SIP Authentication Settings.



The username that was entered in the previous screen will appear in the Private ID field of the
Authentication Options screen. The password will be the last four digits of the ID. (In this case,
the Private ID ends with 1234, so the password will be 1234) After the completion of the
authentication configuration, the user should click OK twice to return back to the main Universal
screen window. The phone will then automatically attempt to register with VeriSign’s SIP
server. If the registration is successful the user should see the screen shown in Figure 45.




       AWPR – Automatic Wireless Phone Registration                                            61
                                                                                  December 5, 2005




                   Figure 45. HP iPAQ Universal Phone Successful VoIP Registration.


Once the phone successfully registers, the phone should be working properly. The user can test
the phone by attempting to make a phone call.


4. PSTN/IP Gateway


Although the local PSTN/IP gateway was not used for the project, the team still wrote a
configuration script for the device. The gateway can be configured via a telnet session. The
purpose of the PSTN/IP gateway is to match up IP addresses to a PSTN number and vice versa.
These matches are made by creating dial peers. The first dial-peer which is necessary is the VoIP
dial peer which sends all traffic dialed out on a certain number to a SIP server. An example
configuration script for this dial peer can be viewed below:


       en
       config t
       dial-peer voice 20 voip
       destination pattern 45886XX T



      AWPR – Automatic Wireless Phone Registration                                          62
                                                                                  December 5, 2005


       codec {g711ulaw}
       session transport udp
       session protocol sipv2
       sip-server {ipv4:209.112.113.19}
       session target {sip-server}
       exit


The en command places the switch in enabled mode. Then, the config t command places the
switch in global configuration mode. Once the switch is in this mode, the user can start entering
dial peer information. The dial-peer voice 20 voip command sets up a new VoIP voice dial peer
which is associated with the number 20. The next command, destination pattern 45886XX T sets
up the dial peer to alter any calls that are going out on a 45886XX number. Next, the codec
{g711ulaw} sets up the outgoing call to use the G.711u codec. After that, the session transport
udp command sets up the call to use the UDP transport protocol which is connectionless. Next,
the user must specify for the dial-peer to use SIP version 2 by using the command session
protocol sipv2. The user should then specify which server the call should be sent to. This can be
done by using the sip-server {ipv4:209.112.113.19} command. As a final step in the voice dial-
peer, the user should specify that the target server is a SIP server. This can be done with the
session target {sip-server} command.


After the user has completed the first script, he or she is ready to prepare a script for the other
direction. This script is called a POTS script and it used to match incoming calls with a specific
port. An example of a POTS script can be seen below.


       en
       dial-peer voice 30 pots
       destination-pattern 45886XX T
       port 0/1:23
       exit




       AWPR – Automatic Wireless Phone Registration                                            63
                                                                                December 5, 2005


The dial-peer voice 30 pots line creates a POTS dial peer which is associated with the number 30.
Then, the destination-pattern 45886XX T command specifies that any incoming calls which are
associated with a 4586XXXX number should be matched to the dial peer. The final command
port 0/1:23 specifies that the incoming call should be forwarded to port 0/1:23.

C. Testing Equipment

During the various tests implemented during the AWPR project, several pieces of testing
equipment were necessary. This equipment was used to fully analyze the call registration
process. The different equipment included an IXIA 1600 Traffic Generator, a Berkeley
Varitronics “Grasshopper” 802.11 Tester, an Agilent Voice Quality Tester, an IXIA IxWLAN,
and a Linksys WET Wireless Bridge.


1. IXIA 1600 Traffic Generator


The IXIA 1600 Traffic Generator was used to generate specific amounts of traffic onto the wired
portion of the network. This helped to test how loading the back end of the access port with
varying amounts of traffic affects the voice quality of the iPAQ phone. Traffic was loaded for
the voice quality test during network saturation. This was done by sending set amounts of traffic
to a device located off of the wireless network.


2. Berkeley Varitronics “Grasshopper” 802.11 Tester


The Berkeley Varitronics “Grasshopper” 802.11 Tester was used to determine the signal strength
at any given time. This was used for both the Automatic Call Registration Signal Level Test and
for the Voice Quality Test at various signal strengths. The “Grasshopper” has a receiver that
allows it to see wireless networks. It can then determine the signal strength of a particular
wireless signal. The receiver sensitivity does not allow it to record signals weaker than -89dBm.
For this reason, that strength is the cut off point used during the testing.




       AWPR – Automatic Wireless Phone Registration                                             64
                                                                                December 5, 2005


3. Agilent Voice Quality Tester


The Agilent Voice Quality Tester was used to determine the quality of the received signal during
the Voice Quality Test. With this device, both the sending and receiving phones are hooked up
to the voice quality tester. The computer connected to the tester then sends a predetermined
signal through the sending phone. Then, the signal received at the other phone is sent to the same
computer where its quality is analyzed. It receives a mean opinion score, MOS, based on the
quality. The method of analyzing the signal that is used is based on the Perception Evaluation of
Speech Quality (PESQ) algorithm.


4. IXIA IxWLAN


The IXIA IxWLAN was used to generate specific amounts of traffic onto the wireless portion of
the network. This was done by using an external source of traffic, the IXIA Traffic Generator,
and sending the forwarding the traffic through the IxWLAN device. This helped to test how
loading the front end of the access port with varying amounts of traffic affects the voice quality
of the iPAQ phone. Traffic was loaded for the voice quality test during network saturation.
Unfortunately, due to limitations of this device, only a limited amount of traffic could be
generated. This amount was not enough to show an affect on the voice quality.


5. Linksys WET Wireless Bridge


The Linksys WET Wireless Bridge was used to test the same thing as the IXIA IxWLAN. It was
used to transmit specific amounts of traffic onto the wireless portion of the network. This also
helped to test how loading the front end of the access port with varying amounts of traffic affects
the voice quality of the iPAQ phone. This device was used in attempt to transmit more traffic
across the network than the IxWLAN device allowed. Unfortunately, it still could not transmit
enough traffic to affect the voice quality.




       AWPR – Automatic Wireless Phone Registration                                           65
                                                                               December 5, 2005



VII. Software

In order to successfully complete the project, the team had to use a number of different software
packages. The software packages included Wireless Valley's "SitePlanner", Camtasia Studio, the
PCTEL roaming client, IXIA IxExplorer, Pocket PC Controller, and Passmark WirelessMon. A
brief discussion of each of these software packages can be found below.

A. Wireless Valley’s “Site Planner”


The Wireless Valley’s “Site Planner” was used to model antennas within the test building. The
program allowed the team to import CAD drawings of the test buildings into the program and
place antennas in specific physical locations. The program simulated the 802.11 propagation
throughout the building and allowed the team to add loss for different types of materials (doors,
inner walls, outer walls, and windows). Although the software is based strictly on calculations
and theories it proved to be fairly accurate when it was used to predict the placement of the
antennas throughout the building. When used in the planning stage, the software was able to
provide information such as the received signal level and the signal to noise ratio. This software
allowed the team to instantly visualize the installed network equipment, determine which antenna
to user for the project, and predict how to contain the RF signals within the building as much as
possible. More information about the software can be found at www.wirelessvalley.com.



B. Camtasia Studio 3


Camtasia Studio is suite of video tools which included an audio editor, menu maker, video
player, recorder, and theater, and video capture. The video capturing tool was used in
conjunction with the Pocket PC Controller program to capture portions of the screen which
displayed the iPAQ’s screen and an Ethereal packet capture. The screen capture program
captured video at a frame rate of 15 frames per second. The Camtasia Studio player was also
used to replay the captured video and the timer in the software was used to record changes in the



      AWPR – Automatic Wireless Phone Registration                                          66
                                                                               December 5, 2005


video. More information about this software package can be found at
http://www.techsmith.com/products/studio/default.asp.



C. PCTEL Roaming Client


The PCTEL roaming client was used by the iPAQ in order to allow the phone to operate on either
the VoIP or GSM network. This software makes the decision as to which network the phone
should connect to. Basically, the phone sends out an occasional SIP registration message to the
SIP server and if the SIP server returns an OK message then the phone registers or remains
registered on the VoIP network. Otherwise, if the phone is unable to send a SIP registration
message or if the phone does not receive an OK message from the server after a specified amount
of time, then the phone automatically switches over to GSM mode. More information about this
software can be found at
http://www.pctel.com/product_group_overview_detail.cgi?id_num=5&styleid=2.



D. Pocket PC Controller


The Pocket PC Controller is a program that allows a Pocket PC user to control the handheld
device via a computer’s monitor, mouse and keyboard. This software allowed the team to
visualize a larger screen, enter in data quicker, and capture the screen of the IPAQ within the
Camtasia recording software. More information about this software package can be found at
http://www.soti.net/default.asp?Cmd=Products&SubCmd=PCPro.

E. IXIA IxExplorer


The IXIA Explorer software serves a management interface for the IXIA traffic generator
chassis. This software allows the user to specify the type and the amount of traffic that he or she
would like to generate on the network. Within the software, the user can change the protocol




       AWPR – Automatic Wireless Phone Registration                                         67
                                                                                 December 5, 2005


type of the traffic, the source address, the destination address, and the type of stream (bursty,
continuous, etc.)

F. Passmark WirlessMon


Passmark WirelessMon is a windows program that allows a user to monitor the status of an
802.11 wireless adapters and gather statistics of surrounding access points. The program
contains capabilities such as signal strength monitoring, multiple access point monitoring,
interference locating, verify security settings of local access points, and aids in checking
coverage and range. The program also has a logging feature that graphs the signal strength and
other parameters over time. More information about this software package can be found at
http://www.passmark.com/products/wirelessmonitor.htm.




       AWPR – Automatic Wireless Phone Registration                                            68
                                                                               December 5, 2005



VIII. Antenna System Analysis
A. Software Validation

1. Introduction


Before using the "SitePlanner" software to predict the in-building coverage provided by various
802.11 antennas, the software was validated in a number of ways. First, the software was used to
model the propagation of an antenna’s signal while in free space. This was done to determine
whether or not the software package gives an accurate representation of received signal strengths
with respect to how far the receiver is from the antenna. This type of test has been performed on
two different types of antennas within the "SitePlanner" software. First, an isotropic antenna was
tested and then a directional cardioid antenna was tested.


After the two types of antennas were tested in the software package, the 802.11 propagation
throughout a building was tested. This process was started by first locating two access points
inside of the test building. Once these access points were located, the transmission parameters
for each of the access points were researched so that the access points could be modeled in
Wireless Valley's "SitePlanner". Next, the group took four signal measurements from each
access point at various locations throughout the building using an 802.11 receiver test device
known as a “Grasshopper”. After this was completed, the access points, along with all of their
transmission parameters, were entered into the "SitePlanner" software. Upon initial simulation
of the signal propagation throughout the building, the software returned a number of false values.
Therefore, the software was adjusted until the software predicted the signal strength at the
measured locations within 5 dB of the actual measured value. Adjustments were made to all of
the following parameters within "SitePlanner": the propagation model, the loss of various
materials, and the loss caused by distance.




       AWPR – Automatic Wireless Phone Registration                                            69
                                                                               December 5, 2005


    2. Procedure

2.1 Isotropic Antenna in Free Space
As a first step towards validating the software, a number of calculations were performed
regarding an isotropic antenna in free space. Here is the formula for an isotropic antenna in free
space:


                     ⎛ λ ⎞
10 log P r = 20 log( ⎜          ⎟ + 10 log Pt
                     ⎝ 4 ⋅π ⋅ r ⎠


For the simplicity of this example, we will assume that the transmission power of the antenna is 1
mW which is equivalent to 0 dBm. This simplifies the equation to the following:


                   ⎛ λ ⎞
10 log Pr = 20 log(⎜          ⎟
                   ⎝ 4 ⋅π ⋅ r ⎠


Next, the formula should be used to find out the radius of an isotropic antenna at various receive
signal levels. By using the logarithmic rules the equation can be changed into the equation seen
below.


    ⎛      λ         ⎞
r = ⎜ (Pr/ 20 )      ⎟
    ⎝ 10        * 4π ⎠


The only unknown variable in this equation is the wavelength, or λ . Lambda can be calculated
by using the following formula:


     ⎛C ⎞    ⎛ 3 X 108 m / s ⎞
λ =⎜ ⎟=⎜
   ⎜ f ⎟ ⎜ 2.437 X 109 Hz ⎟ = 0.123102m
                          ⎟
   ⎝ ⎠ ⎝                  ⎠




         AWPR – Automatic Wireless Phone Registration                                       70
                                                                                December 5, 2005


Note that the frequency used in this formula is 2.437, which is the frequency of channel 6 in
802.11b. Once λ is known, the equation can be simplified to:


    ⎛ 0.123102m ⎞
r = ⎜ (Pr/ 20 )      ⎟
    ⎝ 10        * 4π ⎠


By using this simplified equation, various values of received power (Pr) can be entered to find
out the corresponding radius. For this system, it is ideal to know the distance of the radius
between -50 dBm and -90 dBm. If these values are entered into the formula, the results seen in
Table 4 can be obtained.


                              Table 4. Calculated Isotropic Radius Values.

                           Pr Value      Resulting Isotropic Radius
                           -50 dBm                    3.0878 m
                           -60 dBm                   9.796146 m
                           -70 dBm                    30.978 m
                           -80 dBm                  97.961459 m
                           -90 dBm                  309.78133 m


Using the Wireless Valley “Site Planner”, the user needs to validate the software by simulating
an isotropic antenna in free space. To simulate a free space environment, the antenna should be
placed a significant distance above the ground. This will help reduce any changes in the
radiation pattern which may occur due to the reflection or absorption caused by the ground plane.
Free space should be simulated in Wireless Valley's "SitePlanner" by placing both the antenna
and the receiver 50 feet above the ground and by placing no walls within the building layout.
After this has been completed, the antenna the propagation from the antenna can be simulated
and the contour patterns for various signal levels will appear. The distance tool can then be used
to find out how much distance is between the antenna and the received signal level contour
pattern.




       AWPR – Automatic Wireless Phone Registration                                         71
                                                                                December 5, 2005


In Figure 46, one can see the simulated signal strength levels for -50 dBm (red) and -60 dBm
(yellow). By using the distance tool in "SitePlanner", as seen in Figure 46, the user can
determine that the radius of the isotropic antenna is 2.8 m when at a received signal strength of -
50 dBm.




                           Figure 46. Simulated Isotropic Radius at -50 dBm.


Next, the user can use the distance tool to measure the isotropic radius at -60 dBm (yellow). As
seen in Figure 47, the isotropic radius is 9.63 meters at -60 dBm.




       AWPR – Automatic Wireless Phone Registration                                         72
                                                                              December 5, 2005




                          Figure 47. Simulated Isotropic Radius at -60 dBm.


This process can be continued for all received signal strength levels. Figure 48 shows that the
simulated isotropic radius at -70 dBm (green) is 30.67 meters.




                          Figure 48. Simulated Isotropic Radius at -70 dBm.


Figure 49 below show that the "SitePlanner" predicts a 96.84 meter radius at -80 dBm (blue).




      AWPR – Automatic Wireless Phone Registration                                         73
                                                                                December 5, 2005




                           Figure 49. Simulated Isotropic Radius at -80 dBm.


The final simulated received signal level value, -90 dBm (purple), shows that the radius of an
isotropic antenna at that level should be 306.58 meters as seen in Figure 50.




                           Figure 50. Simulated Isotropic Radius at -90 dBm


The values for the calculated isotropic radius and the software predicted radius at all the received
signal strength levels can be found in Table 5.




       AWPR – Automatic Wireless Phone Registration                                          74
                                                                                     December 5, 2005


           Table 5 . Calculated and Software-Prediction Radii at Various Received Signal Levels.
Pr Value      Calculated Isotropic           “SitePlanner” Predicted              Percentage Difference
                     Radius                       Isotropic Radius               between calculated and
                                                                                software-predicted values
-50 dBm            3.0878 m                            2.80 m                             - 9.3 %
-60 dBm           9.796146 m                           9.63 m                             - 1.7 %
-70 dBm            30.978 m                           30.61 m                            - 1.19 %
-80 dBm          97.961459 m                          96.84 m                            - 1.14 %
-90 dBm          309.78133 m                         306.58 m                            - 1.03 %


 By looking at the table above, one can clearly see that the software accurately simulates an
 antenna in free space.




       AWPR – Automatic Wireless Phone Registration                                                 75
                                                                                December 5, 2005



2.2 Cardioid Antenna in Free Space


Next, for further validation of the software, a directed antenna must be tested in the software.
For the purpose of this validation, a 10 dB cardioid antenna will be used. The distance for the
antenna’s propagation can be calculated by using the following formula:


Pr = Pt + Gt − LFS + Gr


For simplicity, the gain of the receiver (Gr) and the transmitted power (Pr) will be set to 0. The
gain of the cardioid antenna (Gt) should be 10 dB. This will allow us to simplify the formula into
the following:


Pr = 10 − LFS


The formula for free space loss (LFS) is:


LFS = 20 log Dmi + 20 log f GHz + 96.6


Distance in miles is the value that we are looking for and the frequency in GHz is equal to 2.437
GHz. This allows the free space formula to be simplified into the following formula:


LFS = 20 log Dmi + 104.337


By substituting the formula for free space loss into the original formula, the following formula
can be formed:


Pr = 10 − (20 log Dmi + 104.337)




       AWPR – Automatic Wireless Phone Registration                                          76
                                                                               December 5, 2005


Since the distance in miles is the desired value, the formula can be rearranged to appear similar to
the following:


⎛ Pr + 94.337 ⎞
⎜             ⎟ = Dmi
⎝     − 20    ⎠


By using this formula, various received power values can be entered in order to find the range of
the cardioid antenna. After plugging in values between -50 and -90 dBm, Table 6 could be
formed.


                              Table 6. Calculated Cardioid Range Values.

                         Pr Value       Calculated Antenna Range
                          -50 dBm                     9.7678 m
                          -60 dBm                   30.88865 m
                          -70 dBm                     97.678 m
                          -80 dBm                   308.8864 m
                          -90 dBm                     976.784


After the values for the Cardioid antenna have been calculated, the antenna can be entered into
the "SitePlanner" software. Using methods, similar to those used when simulating the isotropic
antenna, the distances for the Cardioid can be found. By looking at Figure 46, one can see that
the cardioid range at -50 dBm (red) is 9.75 meters.




                           Figure 51. Simulated Cardioid Range at -50 dBm.




       AWPR – Automatic Wireless Phone Registration                                         77
                                                                                  December 5, 2005


Next, at -60 dBm (yellow) the simulated distance is 30.48 meters as seen in Figure 52.




                            Figure 52. Simulated Cardioid Range at -60 dBm.


Third, the user can determine the distance of the cardioid at -70 dBm (green). As seen in Figure
53, the software predicts that the range at this received signal level is 97.3 meters.




       AWPR – Automatic Wireless Phone Registration                                         78
                                                                                December 5, 2005




                           Figure 53. Simulated Cardioid Range at -70 dBm.


Next, the user can use the software to predict that the range of the cardioid at -80 dBm (blue) is
305.33 meters. This can be seen in Figure 54.




                           Figure 54. Simulated Cardioid Range at -80 dBm.


As the final step of the cardioid validation process, the user can determine that the simulated
range of the antenna at -90 dBm (purple) is 969.63 meters. This can be seen in Figure 55.




       AWPR – Automatic Wireless Phone Registration                                         79
                                                                                     December 5, 2005




                             Figure 55. Simulated Cardioid Range at -90 dBm.


After the range for -90 dBm has been simulated, the following table can be formed.


            Table 7. Calculated and Software-Predicted Ranges at Various Received Signal Levels.
 Pr Value      Calculated Isotropic Radius        “SitePlanner” Predicted          Difference between
                                                      Isotropic Radius                calculated and
                                                                                   software-predicted
                                                                                          values
 -50 dBm                9.7678 m                           9.75 m                       - 1.822 %
 -60 dBm               30.88865 m                         30.48 m                        - 1.32 %
 -70 dBm                97.678 m                           97.3 m                       - 0.387 %
 -80 dBm               308.8864 m                         305.33 m                       - 1.15 %
 -90 dBm                 976.784                          969.63 m                      - 0.732 %


Like the results seen in Table 5, the results in Table 7 show that "SitePlanner" provides a very
accurate simulated propagation distance in free space.


3. 802.11 Propagation Test


In order to validate the 802.11 propagation throughout the building, the RSSI values received
from two access points were used. A calibration of the software’s building parameters to reflect
the true properties of the Research Park building was completed also using the RSSI values of the



      AWPR – Automatic Wireless Phone Registration                                                  80
                                                                                December 5, 2005


two existing access points. The access points were modeled in "SitePlanner" software and placed
at their approximate location on the floor plan.




                         Figure 56. Modeled Base Stations and Their Coverage.


The software was set through a series of dialog boxes to create a prediction of the RSSI values in
respect to an access point at any particular location; about 20+ values per AP. The values were
recorded as the predicted RSSI values for each AP. The measured RSSI values were recorded
using the “Grasshopper” 2.4GHz WLAN Receiver. The “Grasshopper” was taken to the same
locations inside the building and the RSSI values were measured. Comparison of the measured
and recorded RSSI values showed that the RSSI values inside the building were close to each
other but not close enough. To decrease the difference between the values, the absorption rating
of the inner walls, outer walls, doors, and glass windows were changed using the Partition
Library and the Edit Partition Category (shown in the figure below). After changing and re-
predicting the RSSI values a few times, the software’s building parameters were calibrated to
reflect the true properties of the Research Park building.




       AWPR – Automatic Wireless Phone Registration                                          81
                                                                               December 5, 2005




                              Figure 57. Edit Dialog Boxes for Partitions.


It was found that the software does not predict reliable RSSI values outside of the building. This
is due to the software inability to model outdoors conditions. This is due to the software’s
assumption that anything outside of a building’s walls is theoretically free space. Therefore,
when the signal goes beyond the exterior walls of the building, the only factors affecting the
signal propagation are free space loss and the loss caused by the ground plane. The tables and
figures below display a subset of the predicted and measured values of two of the access points,
Star Vision (SV) and AdvantGX (AGX).




      AWPR – Automatic Wireless Phone Registration                                             82
                                                                             December 5, 2005




       Star Vision
      RSSI (dBm)
 Predicted     Measured
   -82.5         -85
   -74.3         -73
   -65.2         -62
   -64.7         -69
Table 8. Star Vision’s
Measured vs. Predicted
RSSI values.




                               Figure 58. Star Vision's Predicted RSSI Values.




           AGX
      RSSI (dBm)
 Predicted     Measured
   -80.4         -80
   -62.9         -63
   -49.6         -50
   -43.2         -42
Table 9. AdvantGX's
Measured vs. Predicted
RSSI values.




                                Figure 59. AdvantGX's Predicted RSSI Values.



       AWPR – Automatic Wireless Phone Registration                                    83
                                                                                 December 5, 2005




Therefore, calibration of "SitePlanner" was achieved by importing existing access points into the
software into the layout of the first floor of the Texas A&M half of the building in Research
Park. The software was then used to predict the RSSI values from the associated access points at
different locations on the floor plan. The “Grasshopper” 2.4GHz WLAN Receiver was then used
to measure the actual RSSI values from the associated access points. Software parameters were
adjusted until the measured and software predicted RSSI values had the smallest deviation
between them.



B. Antenna Selection


Once the walls in the building were adjusted in order to predict realistic signal levels at any point
in the building, the project team was able to simulate two antennas in the software in order to aid
in the antenna selection process. These two antennas were chosen out of an initial selection of
ten antennas due to radiation patterns. The two antennas that were simulated included the Wi-
Sys Communications 8dBi Waveguide and the Trendnet 6dBi Patch antenna. A more detailed
discussion of these antennas can be found in the following sections.




       AWPR – Automatic Wireless Phone Registration                                          84
                                                                                 December 5, 2005


   1. Wi-Sys Communications 802.11b/g Indoor/Outdoor Waveguide


The first antenna under test was the Wi-Sys Communications 8dBi Waveguide. Figure 60 shows
a picture of this specific antenna.




                              Figure 60. Wi-Sys 8dBi Waveguide Antenna.


The radiation patterns for this antenna can be found in Figure 61 and Figure 62.




      Figure 61. Waveguide Horizontal Radiation
                                                       Figure 62. Waveguide Vertical Radiation
                       Pattern.
                                                                      Pattern.




       AWPR – Automatic Wireless Phone Registration                                              85
                                                                                   December 5, 2005


Table 10 contains the detail specifications for the waveguide antenna.


                             Table 10. Waveguide Detailed Specifications.
                  VSWR                                             Less than 1.4
                  Gain                                             8dBi
                  Polarization                                     Linear
                  Azimuth 3dB BW                                   40 degrees
                  Elevations Plan (3dB BW)                         40 degrees


By knowing these values, the antenna was able to be simulated in "SitePlanner" software. Figure
63 below shows the simulated antenna layout of the first floor. These simulations were
performed with 3 dB attenuators attached to the antennas. The green line indicates the -75 dBm
antenna strength for the antenna placed in the SVT area (upper left), and the red line indicates the
-75 dBm antenna strength for the antenna placed in the ITEC conference room. In this layout,
the upper left antenna would use channel 3 and the bottom left antenna would use channel 9.




       AWPR – Automatic Wireless Phone Registration                                          86
                                                                     December 5, 2005




               Figure 63. Waveguide Antenna Layout on First Floor.




AWPR – Automatic Wireless Phone Registration                                   87
                                                                                 December 5, 2005


Figure 64 contains the waveguide antenna layout for the second floor. For this floor plan, there
will be a single antenna placed in the upper left portion of the floor plan and it will need to be
attenuated by 8 dB.




                        Figure 64. Waveguide Antenna Layout on Second Floor.




       AWPR – Automatic Wireless Phone Registration                                           88
                                                                                      December 5, 2005


   2. Trendnet 6dBi Indoor Directional Patch Antenna


The second antenna under test was the Trendnet 6dBi Patch Antenna. Figure 65 shows a picture
of this specific antenna.




                                  Figure 65. Trendnet 6dBi Patch Antenna.


This antenna was chosen because if its desirable antenna patterns. The horizontal and vertical
patterns can be found in Figure 66 and Figure 67, respectively.




          Figure 66 - Patch Horizontal Pattern                 Figure 67 - Patch Vertical Pattern



Next, the detailed antenna specifications for the patch antenna can be seen in Table 11.


                                  Table 11. Detailed Antenna Specifications.
                   VSWR                                                1.5:1 Max
                   Gain                                                6 dBi
                   Polarization                                        Linear, Vertical
                   Azimuth 3dB BW                                      80 degrees
                   Elevations Plan (3dB BW)                            80 degrees
                   Front to Back Ratio                                 12 dBi



       AWPR – Automatic Wireless Phone Registration                                                 89
                                                                              December 5, 2005




After the radiation patterns and detailed antenna specifications are known, the patch antenna can
be simulated in "SitePlanner" Figure 68 below shows the simulated antenna layout of the first
floor with the patch antennas. The green line indicates the -75 dBm antenna strength for the
antenna placed in the SVT area (upper left), and the red line indicates the -75 dBm antenna
strength for the antenna placed in the ITEC student worker room. In this layout, the upper left
antenna would use channel 3 and the bottom left antenna would use channel 9.




                           Figure 68. Patch Antenna Layout on First Floor.




      AWPR – Automatic Wireless Phone Registration                                        90
                                                                                 December 5, 2005


Figure 69 contains the patch antenna layout for the second floor. For this floor plan, there will be
a single antenna placed in the upper left portion of the floor plan and it will need to be attenuated
by 4 dB.




                           Figure 69. Patch Antenna Layout on Second Floor.




       AWPR – Automatic Wireless Phone Registration                                           91
                                                                                December 5, 2005



IX. Testing
A. Testing Overview
During the AWPR project, five main tests were implemented to verify the functionality of the
network and registration process. Through these tests, the registration process could be fully
analyzed. These tests include Network Association Timing, Automatic Call Registration Signal
Level, Voice Quality, Battery Life, and Antenna Coverage and Containment.

B. Network Association Timing
The network association timing test was used to calculate the registration time. This is the time it
takes to register with one network after deregistering with the other. This test was performed
both from the VoIP to GSM network and from GSM to VoIP. In order to get a close
approximate value for the registration time, an average of 15 samples were taken. The actual test
procedure used for this can be found in part 3 of Appendix A. The average registration time for
VoIP was 1.6425 seconds. The shortest registration time was 0.1168 seconds and the longest
was 2.2051 seconds. Overall, a standard deviation of .4514 was found within the 15 test results.
The test results can be seen in Table 12. The average registration time for GSM is 12.88 seconds.
The lowest value was 11.70 and the highest value was 17.93 seconds. The results contained a
standard deviation of 1.42. The test results can be seen in Table 13. The difference in
registration time between associating with VoIP or GSM is due the iPAQ turning the GSM radio
off while on VoIP.




       AWPR – Automatic Wireless Phone Registration                                         92
                                                                                  December 5, 2005


    Table 12. VoIP to GSM Automatic Registration       Table 13. GSM to VoIP Automatic Registration
                     Test Results.                                     Test Results.
  SIP Registration        SIP OK         Delta              Start        Stop       Delta (seconds)
        Sent             Received      (seconds)             ….            .
         65.388765        67.025844       1.637079          8.51         20.74           12.23
         119.75373       121.205002       1.451272         62.98         75.25           12.27
       165.158772        167.289214       2.130442         108.52        120.42          11.9
       218.241397         218.35821       0.116813         160.95        173.99          13.04
       263.707968        265.913051       2.205083         207.35        219.05          11.7
       314.734852        316.385104       1.650252         256.95        270.26          13.31
       364.752243        366.313677       1.561434         307.48        320.21          12.73
       157.005127        158.769147        1.76402         13.98         26.29           12.31
         70.573343        72.365896       1.792553          5.14         18.07           12.93
       123.353939        125.058333       1.704394         51.54         69.47           17.93
       185.219977        187.018514       1.798537         120.36        132.63          12.27
       241.856385        243.511062       1.654677         176.97        189.2           12.23
       291.827685        293.595313       1.767628         226.56        239.46          12.9
       336.778031         338.52081       1.742779         271.79        284.21          12.42
       379.182353        380.975035       1.792682         314.19        326.85          12.66
       429.495617        431.005519       1.509902         360.74        373.8           13.06
               Average                1.642471688              Average                 12.868125
         Standard Deviation           0.451441884         Standard Deviation            1.42116




Variations in the VoIP registration times are due to the fact that the registration messages must
travel to VeriSign’s SIP server through the VPN connection which has been established over the
commodity Internet. Messages traveling over the commodity Internet are not guaranteed a
committed information rate and their delivery time depends greatly on the traffic loads present on
the paths that they take. The variation found in these times can be due to a number of factors.
The cellular environment works differently based on the number of users connected to the base
station as well as the signal strength of the GSM signal received by the phone. The test plan and
a more detailed discussion of the results can be found in part 4 of Appendix A.




       AWPR – Automatic Wireless Phone Registration                                             93
                                                                                  December 5, 2005


C. Automatic Call Registration Signal Level
The automatic call registration signal level test was used to determine what signal level triggered
the iPAQ to transfer its registration and switch networks. This test was repeated 10 times for
accuracy. It was completed for leaving the VoIP network in idle and dedicated mode, as well as
for leaving the GSM network. The detailed test procedure for this test can be found in the third
part of Appendix B. The average value recorded when leaving VoIP in dedicated mode was
-87.6 dBm. The high value was -89 and the low was -86 dBm. The average value recorded
when in idle mode was -88.4 dBm. The high value was -89 and the low was -88 dBm. The
average value recorded for leaving GSM mode was -86.6 dBm. The high value was -88 and the
low was -84 dBm. The results for each test can be seen below in Table 14. The test plan and a
more detailed discussion of the results can be found in Appendix B.


                                Table 14. Signal Strength Test Results.
Trial #                          1      2       3      4      5       6     7     8     9     10
Signal Strength
                                 -86    -88     -87    -88    -86     -89   -88   -87   -88   -89
(VoIP to GSM Dedicated)
Signal Strength
                                 -89    -88     -88    -88    -89     -89   -88   -89   -88   -88
(VoIP to GSM Idle Mode)
Signal Strength
                                 -88    -86     -84    -88    -86     -88   -86   -88   -87   -85
(GSM to VoIP Idle Mode)




      AWPR – Automatic Wireless Phone Registration                                            94
                                                                                                  December 5, 2005


D. Voice Quality

1. Various Received Signal Strengths


The Voice Quality Test with varying signal strengths was performed to determine how the
received signal strength affects the voice quality. This test was completed from PSTN to VoIP
and from VoIP to PSTN. The results can be seen in the graph in Figure 70.


As one can clearly see, the MOS scores for the PSTN to VoIP voice quality test remained stable
around 3.4 from the -40 dBm level until the -70 dBm. After reaching the -70 dBm received
signal strength level, the voice quality began to degrade. At -80 dBm, the MOS value dropped to
3.1 and then at -90 dBm the voice quality dropped to 2.9. In the test of VoIP to PSTN the results
remained stable regardless of the received signal strength level. This stability could be due to the
fact the voice quality of the phone was poor even at the best possible signal strength, so it could
not degrade much. One reason that the voice quality was degraded so much for these tests was
because of grounding problems that existed between the iPAQ and the VQT. The grounding
problems caused a minor feedback hum to be present in the audio stream coming from the iPAQ,
which lowered the MOS scores. Although extensive research was performed in an attempt to
reduce the feedback, no resolutions were found for this problem. The procedure and results used
to implement this test can be found in part 1 of Appendix C.



                                     MOS vs. Received Signal Strength
       (Mean Opinion Score)




                              4.50
                              4.00
                              3.50
              MOS




                              3.00                                                                PSTN -> VoIP
                              2.50                                                                VoIP -> PSTN
                              2.00
                              1.50
                              1.00
                                  -100   -90   -80     -70    -60    -50    -40     -30
                                                     RSSI (dBm)

                                    Figure 70. MOS vs Received Signal Strength for PSTN & VoIP.



       AWPR – Automatic Wireless Phone Registration                                                         95
                                                                                                 December 5, 2005


   2. Wired Traffic Loading


The Voice Quality Test with wired traffic loading was completed to determine how traffic on the
wired portion of the network will affect the voice quality of a call. This test was performed from
PSTN to VoIP and from VoIP to PSTN. The test was done by loading the back side of the access
point. The results can be seen below in Figure 71. One can see in the results the quality is better
going from the PSTN to VoIP. The results for PSTN to VoIP and VoIP to PSTN intersect at
around 0.8 Mbps. The poor quality for VoIP to PSTN is due to a slight hum which is caused by
grounding issues in the Voice Quality tester and the use of an external microphone jack on the
iPAQ. Further more the VoIP to PSTN could be poorer quality when compared to PSTN to VoIP
because the sound is originating wirelessly and starts of poor. With the PSTN to VoIP the sound
originates on the PSTN which has a guaranteed voice quality level. The procedure and results
for this test can be found in the second part of Appendix C.

                                                          MOS vs. Traffic Load


                               4.25
        (Mean Opinion Score)




                               3.75
                               3.25
               MOS




                               2.75                                                               PSTN -> VoIP
                               2.25                                                               VoIP -> PSTN
                               1.75
                               1.25
                               0.75
                                      0   0.25      0.5    0.75     1     1.25     1.5    1.75
                                             Amount of Traffic Generated (Mbps)

                                                 Figure 71. MOS vs. Traffic for PSTN & VoIP.


   3.        Wireless Traffic Loading


The Voice Quality Test with wireless traffic loading was completed to determine how traffic on
the wireless portion of the network will affect the voice quality of a call. This test was performed
from PSTN to VoIP and from VoIP to PSTN. The procedure and results for this test can be
found in the third part of Appendix C.



        AWPR – Automatic Wireless Phone Registration                                                        96
                                                                                  December 5, 2005


E. Battery Life

The purpose of the battery life test was to determine how long the iPAQ phone could last on
battery power under different circumstances. Three different sets of test were performed for both
idle and dedicated modes. These tests include VoIP mode, GSM mode with the 802.11 Radio
turned on, and GSM mode with the 802.11 Radio turned off. This test was performed both with
the radio on and with it off because in a seamless network implementation, both radios would
have to remain on. However, the ideal battery life occurs when the radio is off. The best
scenario would be if the 802.11 radio had a sleep mode similar to the GSM radio. This would
allow it to send out signals occasionally rather than continuously remaining active on the
network. The test results and procedures used can be found in parts 3 and 4 of Appendix D. The
average battery life time can be seen in Figure 26. The idle mode results from this test are a
VoIP time of 4 hours 18 minutes, a GSM time with the 802.11 radio of 4 hours 15 minutes, and a
GSM time with no radio of 17 hours 13 minutes. The dedicated mode results from this test are a
VoIP time of 3 hours 58 minutes, a GSM time with the 802.11 radio of 3 hours 3 minutes, and a
GSM time with no radio of 7 hours 53 minutes. More detailed test results can be found in parts 5
and 6 of Appendix D.


                                     Table 15. Average Test Results.
                                                    VoIP         VoIP (with
                                        GSM
                                                  (normal)    802.11 radio off)
                         Idle              4:15       4:18         17:13
                         Dedicated         3:03       3:58             9:00



Next, the duty cycles for each the device for each scenario can be calculated. The duty cycle
represents different levels of usage for a user. The duty cycle was calculated A*Y + (1-A)*X
where A is the % of talk time, Y is dedicated talk time, and X is the Idle talk time. The duty cycle
calculation results can be seen in Table 16.




       AWPR – Automatic Wireless Phone Registration                                          97
                                                                                      December 5, 2005


                                          Table 16. Duty Cycles.
                       Talk Time    GSM Duty         VoIP Duty     VoIP (with radio
                      (Dedicated)        Cycle         Cycle       off) Duty Cycle
                          0%             4:15           4:18            17:13
                         10%             4:07           4:16            16:23
                         20%             4:00           4:14            15:34
                         30%             3:53           4:12            14:45
                         40%             3:46           4:10            13:55
                         50%             3:39           4:08            13:06
                         60%             3:31           4:06            12:17
                         70%             3:24           4:04            11:27
                         80%             3:17           4:02            10:38
                         90%             3:10           4:00            9:49
                         100%            3:03           3:58            9:00




C. Antenna Testing


The purpose of this test is to ensure that the antenna system meets the basic requirements stated
in the functional requirements documentation. The main requirements are that the antenna
system should provide at least a -75 dBm signal in 95% of the coverage area and should not
exceed -89 dBm 10 feet outside of the building. With these requirements in place, the 6 dBi
Trendnet patch antenna was chosen. After the access points were placed signal strength of the
access points were measured throughout the building. Based on the measured samples of
received signal strength, the simulations proved to provide fairly accurate received signal levels
through out the building. Only minor adjustment to the location of the access points had to be
made in order to achieve the desired results. Ninety-five percent of the 15582.5 square foot
building is 14,802.9, and with the antennas in place, approximately 14,465.5 square feet or 93%
of the building contained coverage -75 dBm or better. The remaining portions of the building
had a received signal strength between -75 and -85 dBm. These weaker coverage areas were
consistent with the simulated results.




       AWPR – Automatic Wireless Phone Registration                                             98
                                                                                December 5, 2005


In order to test the deployed antenna system, a laptop with the WirelessMon software package
was used to take the measurements. A large number of measurements were taken for each of the
access points. For simplicity, a three-color scale was used to indicate the signal strength seen by
the grasshopper at various spots throughout the building. Green was used to indicate a received
signal strength that was equal to or above -75 dBm. Yellow was used to indicate a signal
strength level between -75 dBm and -85 dBm. Red was used to indicate spots on the map where
the laptop saw a signal strength that was below -85 dBm.


First, the test was performed for the access point operating on channel 9. This access point is
located between in the hallway between the break room and the telephone closet. The access
point and the received signal strength levels for the first floor can be seen on the map shown in
Figure 72.




       AWPR – Automatic Wireless Phone Registration                                         99
                                                                                      December 5, 2005




                      Figure 72. Channel 9 Access Point’s Radiation on First Floor.


As shown in Figure 72, the received signal strength for the channel 9 access point almost meets
all of the requirements. All of the received signal strengths 10 ft outside of the building are -80
dBm or less. In most cases the signal strength at 10ft outside of the building is -89 dBm. In
addition, this antenna provides a received signal strength level of -75 dBm and greater almost
everywhere on the first floor. Next, the access point’s signal strength was measured on the
second floor. These measurements can be seen in Figure 73.


       AWPR – Automatic Wireless Phone Registration                                             100
                                                                           December 5, 2005




          Figure 73. Channel 9 Access Point’s Radiation on the Second Floor.




AWPR – Automatic Wireless Phone Registration                                         101
                                                                                    December 5, 2005


Next, the access point located in the SVT area on the first floor was analyzed. This access point
operates on channel 3. The placement of the access point as well as the received signal strength
levels for the access point on the first floor can be seen in Figure 74.




                  Figure 74. Channel 3 SVT Access Point’s Radiation on the First Floor.




       AWPR – Automatic Wireless Phone Registration                                           102
                                                                                  December 5, 2005


Next, the signal strength was measured for the first floor SVT access point on the second floor.
The results from these measurements can be seen in Figure 75.




                Figure 75. Channel 3 SVT Access Point’s Radiation on the Second Floor.



      AWPR – Automatic Wireless Phone Registration                                          103
                                                                                     December 5, 2005




The final access point is located on the second floor and it also operates on channel 3. The first
floor signal measurements for this access point can be seen in Figure 76.




                Figure 76. Channel 3 Upstairs Access Point’s Radiation on the First Floor.


The second floor received signal strength measurements for the second floor access point can be
seen in Figure 77.




       AWPR – Automatic Wireless Phone Registration                                            104
                                                                           December 5, 2005




      Figure 77. Channel 3 Upstairs Access Point’s Radiation on the Second Floor.




AWPR – Automatic Wireless Phone Registration                                         105
                                                                                December 5, 2005



X. Conclusions

After completing this project, the members in the Ringer Communications team have learned an
extensive amount of knowledge regarding VoIP systems including device registration and call
flow. The project team also learned one potential method that may be used in the future to
integrate cellular and VoIP networks. Although a number of obstacles still remain in the way for
a seamless VoIP and GSM integration, the team has analyzed the situation and has provided
suggestions for the VeriSign design team which may be considered in the future.


In addition, the team has learned how to perform a number of network tests. The skills learned in
this area include voice quality testing, traffic loading, signal strength measurement. While
performing the voice quality testing research the team learned how voice quality scores are
assigned to calls and the factors which affect these scores. After implementing the voice quality
tests, the group was able to use this research to analyze the data that was collected. By
researching and completing the traffic loading tests, the project team was able to become familiar
with the network traffic generation equipment including the IXIA traffic generator chassis and
the IxWLAN wireless tool. The signal strength measurement tests helped the team learn the
propagation of RF signals throughout a building and aided in the team’s efforts to maximum
coverage and containment within the building.


In addition to the skills already mentioned, the project team has also learned a large amount of
knowledge about antenna systems. First, the group learned how to model an antenna within the
"SitePlanner" software and how to validate the software using formulas. Then, the group learned
the skills necessary to install the antenna system and optimize the antenna system for maximum
coverage within the building. This software proved to be rather useful in planning the
deployment of the antennas throughout the building and making an antenna selection.


Although there is still a large amount of research to be performed in the VoIP and GSM
convergence arena, it is the belief of the Ringer Communications team that the test results from
this project can be used to help A&M make one more step towards the future. Hopefully, the test



       AWPR – Automatic Wireless Phone Registration                                         106
                                                                            December 5, 2005


results can be used by the Telecommunications department to make a decision on whether or not
a similar campus-wide service should be provided to students and faculty.




      AWPR – Automatic Wireless Phone Registration                                    107
                                                                                 December 5, 2005



XI. Recommendations for Further Research

After completing the research, testing, and implementation of the AWPR project, several new
issues were brought to the project team’s attention. These issues involved several aspects of the
project, and suggestions were made to the necessary entities in an attempt to have these issues
resolved in the future. The project team believes that it would be beneficial to the involved
parties to execute further research and developments within the phone and the call registration
process before implementing a similar service in a production environment.


At this moment in time, a seamless call handoff between the VoIP and GSM network is not
possible. This is due to several factors, some being iPAQ limitations, and others being
limitations of the equipment completing the registration transfer. The further research of
seamless handoffs would not only prove beneficial in gaining a better understanding of the
registration process, but it would also increase the practicality of the network infrastructure as a
whole. Nortel, as well as a number of other companies, are working on a solution known as
pass-through registration in which a seamless call handoff would be possible. This type of
solution would involve a soft handoff between the two networks meaning that once a phone’s
802.11 signal strength level was degraded to a certain point, a server would make a second call
out to the cellular network and perform a handoff to the GSM network only when it became
absolutely necessary. In theory, this type of switchover between networks would allow a user to
roam between GSM and VoIP in mid-conversation without a problem. It is Ringer
Communications’ belief that further research should be performed on the seamless call
registration before any decisions are made regarding whether this service should be deployed
campus-wide.


Another factor that might need to be researched further by the manufacturer of the phone before
implementing the system is the ability to control the transmission power of the phone while
connected to the wireless LAN. While in the cellular mode, the dual-mode phone is able to
adjust its power level as necessary to provide a better connection between the device and the
service provider’s tower. A variable power solution should be implemented in the 802.11 radio



       AWPR – Automatic Wireless Phone Registration                                           108
                                                                               December 5, 2005


system in order to prevent unnecessary power consumption caused by a high transmission power
when a device is close to an access point. In addition, the adaptive 802.11 radio control would
help the conversation voice quality when the user is further away from the access point by
increasing the transmission power.


Next, the phone manufacturer should perform more research in the area of power consumption of
the 802.11 radio. In order to implement a seamless solution, both the 802.11 and GSM radios
must remain on at all times. The GSM radio of the phone has the ability to enter a sleep mode
which means that the radio only sends messages to the provider occasionally rather than
continuously. By utilizing the sleep mode, the phone is able to have a much longer battery life.
Currently, the 802.11 radio in the phone has no means of entering a sleep mode similar to the
GSM radio. As the battery life test results prove, having the 802.11 radio fully enabled all the
time caused the phone’s battery life to be greatly reduced. Further research should be performed
regarding a sleep mode for the 802.11 radio in order to reduce the unnecessary battery
consumption.


Another suggestion for further research that was made by the Ringer Communications team to
VeriSign is that a more complex SIP server should be used for the implementation of the
converged solution. The SIP server used in this project was not able to perform internal SIP calls
meaning that all SIP calls had to leave the network, travel to the PSTN, and re-enter the network.
This long call flow process alone can greatly affect the voice quality of a phone conversation and
therefore a different type of server should be used for this implementation. In addition, the new
SIP server should also support a form of SMS messaging for the phone to use while it is on the
VoIP network. It was found that the phone was not able to send text messages to any other
cellular users while registered on the VoIP network, and would pose a serious problem to users in
a realistic implementation of the service.


The next suggestion given by the project team to VeriSign is that QoS (Quality of Service)
should be used by the phone in order to provide network priority for the voice data coming from
the phone. This will cause VeriSign to either use a different VoIP soft client for the dual-mode
phone or implement QoS within their existing client. QoS is widely supported by most modern-


       AWPR – Automatic Wireless Phone Registration                                          109
                                                                                 December 5, 2005


day networks, but it has to be set by the initiating device. In the case of this network, the
initiating device is the dual-mode phone. The current soft client used by the phone does not set
the QoS, so the voice traffic coming from the phone has the same priority as all the other devices
on the network. This means that voice packets from the phone have the same probability of
being dropped as unimportant Internet or e-mail packets.


The first suggestion given to the ITEC staff by the project team was that a new voice quality
tester be purchased. The current voice quality tester does not always work properly if a device is
not properly grounded to the voice quality tester itself. In the case of testing some devices, such
as the iPAQ it is very difficult to ground the device to the VQT in an attempt to achieve optimum
MOS scores. In addition, to the grounding issues existent with the current VQT, it runs on a
Window 95 operating system while most new VQTs run on a much more up-to-date operating
system.


The final suggestion made by the project team is that if the university chooses to implement the
solution campus-wide then all of the VeriSign equipment should be located locally rather than in
Virginia. By forcing all SIP traffic to travel to Virginia and back, a large amount of latency is
added to the transmission time of every message. Therefore, the SIP server and the Wireless
Mobile Gateway components of the network should both be located on the A&M network in
order to greatly increase the performance of the network and the wireless VoIP service.




       AWPR – Automatic Wireless Phone Registration                                             110
                                                       December 5, 2005



XII. Glossary

AP           Access Point
ACWP         Actual Costs of Work Performed
AWPR         Automatic Wireless Phone Registration
BCWS         Budgeted Costs of Work Scheduled
BCWP         Budgeted Costs of Work Performed
BS           Base Station
BSC          Base Station Controller
BW           Bandwidth
CAD          Computer Aided Drafting
CPI          Cost Performance Index
DHCP         Dynamic Host Configuration Protocol
DTE          Data Terminal Equipment
EVP          Earned Value Process
GSM          Global System for Mobile Communications
HLR          Home Location Register
HP           Hewlett Packard
IP           Internet Protocol
ITEC         Internet 2 Technology Evaluation Center
LAN          Local Area Network
LLC          Logical Link Control
LWAPP        Light Weight Access Point Protocol
MAC          Media Access Control
MOS          Mean Opinion Score
MSC          Mobil Switching Center
NLD          Network Logic Diagram
NRD          Networking Routing Directory
ODC          Other Direct Costs
OSI          Open Systems Interconnect



       AWPR – Automatic Wireless Phone Registration              111
                                                        December 5, 2005


PESQ          Perception Evaluation of Speech Quality
PoE           Power over Ethernet
POTS          Plain Old Telephone Service
PPC           Pocket PC
PSTN          Public Switched Telephone Network
PVID          Port VLAN Identifier
QoS           Quality of Service
RAM           Responsibility Assignment Matrix
RF            Radio Frequency
RSSI          Received Signal Strength Indicator
RTCP          RTP Transmission Control Protocol
RTP           Real-Time Protocol
SDP           Session Description Protocol
SIP           Session Initialization Protocol
SMS           Short Messaging System
SPI           Schedule Performance Index
SVT           Star Vision Technologies
UDP           User Datagram Protocol
VLAN          Virtual Local Area Network
VoIP          Voice over Internet Protocol
VPN           Virtual Private Network
WEP           Wired Equivalent Privacy
WBS           Work Breakdown Structure
WLAN          Wireless Local Area Network
Wi-Fi         Wireless Fidelity




        AWPR – Automatic Wireless Phone Registration              112
                                                  December 5, 2005



XIII. Appendix




      Appendix



   AWPR – Automatic Wireless Phone Registration             113
                                                                                 December 5, 2005



A. Network Association Timing Test

1. Objective


The purpose of this test was to determine the amount of time required for the HP iPAQ 6315 to
automatically register to the desired environment depending on the user’s location. If the user is
within the A&M half of the 1700 Research Park building, the phone will automatically register
with VeriSign’s SIP server through a wireless 802.11b network. Otherwise, if the user is outside
the building, the phone will automatically register with the cellular service provider.


The measurement of the timing when moving from the GSM environment to the 802.11
environment was performed by analyzing the SIP traffic that traveled between the dual-mode
device and VeriSign’s call registration server. Upon moving into the 802.11b coverage area, the
phone will automatically send a registration message to VeriSign’s SIP server and the server will
return an acknowledgement, or an OK message. Based on this information, the time to
automatically register on the VoIP network was defined as the time interval between when the
phone sent an INVITE to the SIP server and when the phone received an OK, or
acknowledgement, from the SIP server to inform the phone that had registered on the network.
After the phone received the OK status message, it could then begin receiving and making calls
on the VoIP network rather than the GSM network.


It is more difficult for the opposite situation to be measured. When the user moved from the
wireless VoIP environment to the GSM environment, the only way to measure the actual
registration time was to monitor the GSM traffic between the phone and the cellular service
provider. Since GSM monitoring equipment was not available to Ringer Communications for the
project, the team had to measure the time it that it took for the phone’s client to switch from VoIP
to GSM mode. This was done by capturing a video from the iPAQ screen using the Pocket PC
Controller software. Since the phone’s client displays the network that the phone is currently
registered to, a video capture was used to estimate how long it took for the soft client to indicate
that the phone had switched from the VoIP to GSM network. The captured video was then
played using Camtasia Studio in order to see specific times at which the phone switched between


       AWPR – Automatic Wireless Phone Registration                                          i
                                                                               December 5, 2005


networks. Although this measurement was not the true automatic registration time, it provided a
good estimate for the time required to complete the switching process.

2. Equipment
The network association timing test involved a number of different software packages and
several pieces of equipment for its execution. The following equipment and software was used
for the testing:


    •   Hardware
            o Laptop running Windows XP
            o Hp iPAQ 6315
            o iPAQ Hot Sync cradle


    •   Software
            o VeriSign Universal phone
            o Pocket PC Controller
            o Camtasia Studio
            o Ethereal


3. Test Plan

3.1 Automatic VoIP Registration


As previously stated in the introduction, the automatic VoIP registration time was measured by
monitoring the packets sent between the phone and the SIP server. These packets were captured
by utilizing the Ethereal software package on a laptop connected to the same wireless network as
the iPAQ.


The VoIP registration time was defined as the time elapsed between the sending of a SIP INVITE
message and the server’s acknowledgement of the registration. To see the VoIP automatic
registration time, the user should begin by opening Ethereal. Once this is done, the user can click



        AWPR – Automatic Wireless Phone Registration                                       ii
                                                                                December 5, 2005


on the Capture Menu and select Interfaces. Next, the user should select the proper wireless
interface and begin capturing packets.


After Ethereal has begun capturing packets, the user should walk outside of the beginning into
the GSM environment. At this point, the user should wait until the soft client on the iPAQ
indicates that it has transitioned to the GSM environment. After the iPAQ has done so, the user
should move back into the wireless VoIP environment. Upon making these changes in location,
the iPAQ will automatically associate with the wireless 802.11b network. The iPAQ will
complete the initial association with the access point, and then it will automatically send a SIP
registration request to VeriSign’s SIP server.


The SIP server will then send a 407 message back to the device which means that proxy
authentication is required. Upon receipt of the 407 message, the phone will send another
registration message to the SIP server which contains the authentication username and password.
After the SIP server receives the second registration, it will return a 200 status message back to
the iPAQ to let the client know that he or she is registered to VeriSign’s VoIP network.
Meanwhile, all of the packets that have been mentioned are captured in Ethereal on the laptop
and can be seen as soon as the user stops capturing packets. An example of an SIP registration
packet capture can be seen in Figure 78. The SIP messages are located within the red rectangle
seen in the figure.




       AWPR – Automatic Wireless Phone Registration                                          iii
                                                                              December 5, 2005




                             Figure 78. SIP Registration Packet Capture.


In the example seen the SIP registration message is sent out at timestamp 370.09495 seconds and
the OK status message is returned at 370.40222 seconds. By taking the difference between these
timestamps, one is able to determine that the registration process takes approximate 0.30727
seconds.



3.2 Automatic GSM Registration


Due to the unavailability of GSM monitoring equipment, a true GSM registration test could not
be performed. Ringer Communications measured the time that it took for the soft client to
inform the user that the phone had switched to the GSM network. This time was measured by
taking a video capture of the iPAQ screen and measuring the interval between when the phone
indicated that it was on the VoIP network and when the phone indicates that it was on the GSM
network.


In order to begin the automatic GSM registration timing test, the user should launch the Pocket
PC Controller program on a laptop which is connected to the iPAQ via a USB cradle. This
program will allow the user to view and record the iPAQ screen on the laptop and save the video
capture as a Window movie file. When the user begins recording, he or she should be located in


      AWPR – Automatic Wireless Phone Registration                                        iv
                                                                                December 5, 2005


the VoIP environment. Next, after recording has begun the user should move outside the
building to the GSM environment. Once the phone indicates that it has switched to the GSM
network, the user can cease recording in the Pocket PC Controller software.


After the file has been saved on the laptop, the user can play the video capture to see how long it
took for the client to indicate a switch from the VoIP network to the GSM network. A program
other than Windows Media Player should be used for the playback since Windows Media Player
only display videos at one second intervals. For the purpose of this test, Camtasia Studio is used
since it allows the user to play back the video at rate up to 15 frames per second. As seen in
Figure 79, the user should step through the video until right before the phone switches off of the
VoIP network. When the user reaches this point, the timestamp should be recorded.




                           Figure 79. Video Capture of Phone in VoIP Mode.


Next, the user should step through the video one interval at a time until the phone indicates that it
has switched to the GSM network as seen in Figure 80. When the user finds this point in the
video, the timestamp should be recorded.


       AWPR – Automatic Wireless Phone Registration                                          v
                                                                                December 5, 2005




                           Figure 80. Video Capture of Phone in GSM Mode.


Then by taking the difference between the first and second timestamps, the user can determine
the time that it took for the client to indicate that the phone has switched from the VoIP to the
GSM network. In this example, it took the phone 12.3 seconds to indicate that it had switched
between networks.




4. Test Results

4.1 Automatic VoIP Registration


First, the test was performed for the automatic VoIP registration process. The results from all 15
trials can be seen in Table 17.




       AWPR – Automatic Wireless Phone Registration                                          vi
                                                                                   December 5, 2005



                      Table 17. VoIP to GSM Automatic Registration Test Results.
                         SIP Registration         SIP OK            Delta
                               Sent              Received         (seconds)
                                 65.388765         67.025844         1.637079
                                 119.75373        121.205002         1.451272
                                165.158772        167.289214         2.130442
                                218.241397         218.35821         0.116813
                                263.707968        265.913051         2.205083
                                314.734852        316.385104         1.650252
                                364.752243        366.313677         1.561434
                                157.005127        158.769147          1.76402
                                 70.573343         72.365896         1.792553
                                123.353939        125.058333         1.704394
                                185.219977        187.018514         1.798537
                                241.856385        243.511062         1.654677
                                291.827685        293.595313         1.767628
                                336.778031         338.52081         1.742779
                                379.182353        380.975035         1.792682
                                429.495617        431.005519         1.509902
                                      Average                    1.642471688
                                 Standard Deviation              0.451441884



The average automatic registration for the GSM to VoIP registration process took 1.64 seconds
and the standard deviation was .45 seconds. Variations in the VoIP registration times are due to
the fact that the registration messages must travel to VeriSign’s SIP server through the VPN
connection which has been established over the commodity Internet. Messages traveling over the
commodity Internet are not guaranteed a committed information rate and their delivery time
depends greatly on the traffic loads present on the paths that they take.




       AWPR – Automatic Wireless Phone Registration                                          vii
                                                                                     December 5, 2005



4.2 Automatic GSM Registration


Next, the automatic GSM registration testing was performed. The results from all 15 trials can
be found in Table 18.


                        Table 18. GSM to VoIP Automatic Registration Test Results.
                                  Start         Stop      Delta (seconds)
                                   8.51         20.74          12.23
                                  62.98         75.25          12.27
                                  108.52        120.42          11.9
                                  160.95        173.99         13.04
                                  207.35        219.05          11.7
                                  256.95        270.26         13.31
                                  307.48        320.21         12.73
                                  13.98         26.29          12.31
                                   5.14         18.07          12.93
                                  51.54         69.47          17.93
                                  120.36        132.63         12.27
                                  176.97        189.2          12.23
                                  226.56        239.46          12.9
                                  271.79        284.21         12.42
                                  314.19        326.85         12.66
                                  360.74        373.8          13.06
                                      Average                12.868125
                                 Standard Deviation           1.42116



The average GSM registration time was found to be 12.87 seconds and the standard deviation
was 1.42116 seconds. The variation found in these times can be due to a number of factors. The
cellular environment works differently based on the number of users connected to the base
station as well as the signal strength of the GSM signal received by the phone.




      AWPR – Automatic Wireless Phone Registration                                             viii
                                                                                December 5, 2005



B. Automatic Call Registration Signal Level

1. Objective


The objective of this test was to identify what signal level caused the HP iPAQ 6315 to switch
between networks. Due to the receiver sensitivity of the iPAQ, the device will never show a
received signal strength less than -90 dBm. Therefore, it is not possible to distinguish a
difference between a received signal strength of -90 dBm and -100 dBm while using the iPAQ.
Because of this, a call can sometimes be made while the iPAQ indicates a -90 dBm signal
strength and at other times a call cannot be completed. In order to overcome this problem, an
802.11b spectrum analyzer called the “Grasshopper” was used to record a more precise received
signal strength level. The “Grasshopper”, like the iPAQ, has a -89 dBm receiver sensitivity but
the “Grasshopper” has a further range due to its additional 3dBi antenna. In addition, the
“Grasshopper” is a test device so it provided more accurate measurements than the iPAQ and it
provided a real-time reading.


This test was conducted in three different sections. The first part of the test was used to determine
the 802.11b signal level at which a VoIP call was dropped and the phone automatically
connected to the GSM network. This was called the VoIP dedicated mode test, since the phone
did not have a call in progress while the test was being performed. The second part of the test
was used to record the signal level at which the phone switched registration from the VoIP
network to the GSM network while in idle mode. Idle mode means that this test was performed
on the phone while an active conservation was not in progress. The third and final test was used
to record what 802.11 signal level caused the iPAQ in idle mode to switch from the GSM
network to the 802.11 network. All of the signal strength measurements that were recorded for
each of these tests were taken from the “Grasshopper.”




       AWPR – Automatic Wireless Phone Registration                                          ix
                                                                                       December 5, 2005


2. Equipment


The automatic call registration signal level testing involved two pieces of hardware and a single
piece of software. These utilized components can be found below:


Hardware
        HP iPAQ 6315
        “Grasshopper” Handheld Spectrum Analyzer
Software
        VeriSign Universal Phone


3. Test Plan

3.1 VoIP to GSM in Dedicated Mode


(3.1.1) While on the 802.11 network, the user should place a VoIP call.
(3.1.2) Then, the user should slowly walk away from the 802.11 coverage area.
(3.1.3) In this case, the user should walk towards the exit of the building and eventually outside of the
        building.
(3.1.4) The user should observe the amount of noise on the line as the signal grows weaker.
(3.1.5) As the noise increases, the user should decrease the speed of walking, ensuring that the phone
        remains connected to the VoIP network.
(3.1.6) Once the call is dropped, the user should mark the location on the ground where this drop
        occurred.
(3.1.7) Next, the user should place the Grasshopper at the location determined in step 3.1.4 and record
         the shown signal level.
(3.1.8) The user should keep the location marked because it will give a rough estimate of where,
        geographically, the other tests might possibly switch between networks.




       AWPR – Automatic Wireless Phone Registration                                                 x
                                                                                       December 5, 2005


3.2 VoIP to GSM in Idle Mode


(3.2.1) The user should begin the test by ensuring that the iPAQ display shows that it is connected to
        VoIP, instead of GSM.
(3.2.2) The user should slowly walk away from the 802.11 access point.
(3.2.3) Then, the user should decrease their speed as they reach the mark that was created in step 3.1.4.
(3.2.4) Once the phone display switches from VoIP to GSM, the user should record this location on the
        ground.
(3.2.5) Next the user should complete steps 3.1.7 and 3.1.8 for the spot marked in the previous step.
(3.2.6) As a final step, the user should compare the signal level at which a call is dropped to the level at
        which the phone display switches networks.

3.3 GSM to VoIP in Idle Mode




(3.3.1) First, the user should walk outside the building and ensure that the iPAQ display
        shows that it is connected to GSM, instead of VoIP.
(3.3.2) Next, the user should slowly walk towards the entrance of the building
        approaching the marks made while performing the previous two tests.
(3.3.3) The user should decrease their speed as they reach these marks.
(3.3.4) Once the phone display switches from GSM to VoIP, the user should record this
        location on the ground.
(3.3.5) As a final step, the user should complete steps 3.1.7 and 3.1.8 for the spot marked in the previous
        step.


4. Test Results

4.1 VoIP to GSM in Dedicated Mode


For this portion of the signal testing, the user measured the signal level at which a VoIP call was
dropped when moving from the VoIP to GSM environment. This test was repeated ten times to
ensure that an accurate cutoff level was recorded. The results from this test can be seen in Table
19.


       AWPR – Automatic Wireless Phone Registration                                                  xi
                                                                                     December 5, 2005




                   Table 19. VoIP to GSM Dedicated Mode Signal Strength Test Results.
Trial #                              1     2     3     4      5      6      7        8     9     10
Signal Strength (dBm)                -86   -88   -87   -88    -86    -89    -88      -87   -88   -89



4.2 VoIP to GSM in Idle Mode


For second portion of the signal testing, the signal level at which the phone’s software switched
from VoIP to GSM in idle mode was measured. While performing this test, it was found that the
phone stayed on the VoIP network a couple of feet further than the previous test. However,
when a call was made to the iPAQ outside of the range seen in the previous test, the call could be
connected, but it was dropped immediately after the connection was made. Similar to the
previous test, 10 samples were taken for this test. The results from this test can be seen in Table
20.


                     Table 20. VoIP to GSM Idle Mode Signal Strength Test Results.
Trial #                              1     2     3     4      5      6      7        8     9     10
Signal Strength (dBm)                -89   -88   -88   -88    -89    -89    -88      -89   -88   -88




4.3   GSM to VoIP in Idle Mode


For the final portion of the signal testing, the signal level at which the phone switches back from
GSM to VoIP was measured. For this test, 10 samples were taken for accuracy. The results from
this test can be seen in Table 21.


                     Table 21. GSM to VoIP Idle Mode Signal Strength Test Results.


Trial #                              1     2     3     4      5      6      7        8     9     10
Signal Strength (dBm)                -88   -86   -84   -88    -86    -88    -86      -88   -87   -85



          AWPR – Automatic Wireless Phone Registration                                           xii
                                                                                December 5, 2005



C. Voice Quality

1. Various Received Signal Strength Levels

1.1 Objective


The main purpose of the various received signal strength level voice quality testing that was
performed in this project was to quantify the voice quality of the HP iPAQ 6315 in a wireless
VoIP network under various received signal strength conditions. The test was also used to
compare the voice quality of the designed wireless VoIP network to the GSM network. The
voice quality in this testing was quantified using Mean Opinion Scores, or MOSs. When MOS
scores were originally created, a panel of expert listeners would rate the voice quality of a phone
conversation, or an audio clip, on a scale between 1 and 5. Using this standard, the toll quality
for a standard landline phone conversation was determined to be between 3.7 and 4. Nowadays,
voice quality test devices are able to assign a voice quality score to a phone call using a more
advanced technique known as PESQ, or Perceptual Evaluation of Speech Quality. This method
matches collected subjective data obtained from thousands of expert listener’s scores to rate a
call objectively. A voice quality testing device also takes packet loss, one-way delay, and jitter
into account when assigning a PESQ score to a call.



1.2 Equipment


The voice quality tests were fairly complex and involved a number of different devices. The
following devices were needed to complete the test:
•      Nortel BayStack 460
•      Nortel WLAN 2230
•      Nortel WLAN Security Switch
•      HP iPAQ 6315
•      Agilent Voice Quality Tester




       AWPR – Automatic Wireless Phone Registration                                          xiii
                                                                               December 5, 2005


•      J1996 VQT Phone adapter – amplifier to help adjusts the audio input and output for each
device under test
•      PSTN Phone

1.3   Test Setups


In order to fully evaluate the voice quality for Ringer’s fully implemented system, three different
scenarios were evaluated. First, the voice quality between a PSTN phone and the iPAQ in
wireless VoIP mode were evaluated. Second, the voice quality between a PSTN phone and the
dual-mode phone in GSM mode were evaluated. As a final step, the voice quality between two
iPAQs in VoIP mode was evaluated. Although the test setup for each of the tests were
essentially the same, the networks to which the devices were connected to were different. In the
two tests involving the wireless VoIP network, the test was performed at various received signal
strengths, which will be discussed later in this document. The three main test setup scenarios can
be seen in the following sections.


       PSTN     VoIP (iPAQ) and VoIP     PSTN


First, a call was placed from the wireless VoIP phone to a PSTN phone. After the call was setup,
the voice quality was evaluated for an audio stream originating from the PSTN phone and
arriving at the iPAQ. Next, the voice quality was evaluated for an audio stream originating from
the iPAQ and arriving at the PSTN phone. The test setup for this scenario can be seen below in
Figure 81.




       AWPR – Automatic Wireless Phone Registration                                         xiv
                                                                             December 5, 2005




                                Figure 81. PSTN <-> VoIP Test Setup.


In the test setup, seen above, the PSTN phone was connected to A&M’s telephone network with
an RJ-11 cable. Then, the RJ-22 handset plug from the phone was connected to the J1996 VQT
phone adapter. Next, the E&M port of the VQT phone adapter was plugged into the C port of the
VQT using a standard RJ-45 cable. The D port of the VQT was then connected to a second VQT
phone adapter. After that, the handset port of this adapter was connected to the iPAQ’s handset
jack using a special RJ-22 to 3.5 mm audio cable.


       GSM (iPAQ)      VoIP(iPAQ) and VoIP (iPAQ)      GSM (iPAQ)


The PSTN to GSM tests was setup identical to the previous test. The only difference between the
two tests was that the phone was set to operate on GSM mode rather than wireless VoIP mode.
The test setup can be seen in Figure 82.




      AWPR – Automatic Wireless Phone Registration                                       xv
                                                                             December 5, 2005




                               Figure 82. PSTN <-> GSM Test Setup.


       VoIP (iPAQ)    VoIP(iPAQ)


The final test setup utilized two iPAQs rather than the PSTN phone and an iPAQ. In this test
setup, both iPAQs used special 3.5mm to RJ-22 cables in order to connect to the VQT adapters.
The VoIP to VoIP test setup can be seen in Figure 83.




                               Figure 83. VoIP <-> VoIP Test Setup.


1.4 Test Procedure




      AWPR – Automatic Wireless Phone Registration                                       xvi
                                                                                December 5, 2005


To begin the test, the two devices to be used in the tests should be connected to the voice quality
adapters. These adapters should each be wired to the E&M ports on the back of the Agilent VQT
(Voice Quality Tester). The adapters are used in order to adjust for discrepancies between the
input and output audio levels of the different devices under test. In order to calibrate the gain on
these boxes for each scenario, the best possible case should be used. For example, when using
the phone on VoIP, the boxes should be calibrated when the phone has a received signal strength
of -40 dBm. This is the most likely the highest signal strength that an iPAQ will receive from an
access point.


To begin the adapter’s gain calibration process, the user should ensure that the Ear and Mouth
gain on each of the two adapter boxes are initialized to zero. Next, the user should launch the
Voice Quality Tester program and begin a Signal Loss Test. This can be done by selecting the
Signal Loss Option from the Measurements menu. Then, the user should set the audio source to
be Port C and the Audio Destination to be Port D. Next, the user should begin the test by
clicking on the Start button. The results should appear similar to those seen in Figure 84.




                                     Figure 84. Signal Loss Test.




       AWPR – Automatic Wireless Phone Registration                                           xvii
                                                                                 December 5, 2005


As seen in the previous figure, the software will generate a graph along with a Mean Signal
Loss/Gain value. This value tells the user how much the gain boxes should be adjusted in order
to compensate for the devices under test. If the value is above the desired signal level, then the
Ear and Mouth amplification settings on the adapter boxes should be increased. If the value is
below the desired signal level, then the settings should be decreased. If the receiving device is
the dual-mode phone then the desired level is 3 dB, but if the receiving device is a wireline VoIP
phone or a PSTN phone, then the desired signal level is 0 dB. These values were found in the
Agilent Voice Quality Tester manual. After the amplification adapters have been adjusted so that
the Signal Loss/Gain level is at the appropriate value, the user should swap the source and
destinations ports and calibrate the voice quality adapters for the reverse direction.


Once both directions have been calibrated, the user should begin the Clarity Trend test. The
Clarity Trend test performs a series of clarity tests which assign an MOS value to the audio
received by the test device. After completing the trend, the program takes an average of all of the
data seen throughout all of the trials. When performing the Clarity Trend test, the user can
specify the number of samples that should be taken. A screenshot of a Clarity Trend test can be
seen in Figure 85.




       AWPR – Automatic Wireless Phone Registration                                           xviii
                                                                               December 5, 2005




                                    Figure 85. Voice Quality Test.


As seen above, the Clarity Trend test will display several calculated values upon completion.
The most important values derived from this test are the average PESQ LQ value and the average
estimated delay. The average PESQ LQ is the listening quality rating of the calls in terms of an
MOS score.


After the test clarity trend has been completed for the best case in each of the two VoIP
scenarios, the voice quality adapters should remain at their set amplifier levels. This should be
done in order to fully represent voice quality changes which occur due to signal loss that occurs
in the audio stream due to weaker received signal strengths. After the baseline is set for wireless
VoIP at a received signal strength of -40 dBm, the user should move one of the iPAQs to one of
the other pre-selected received signal levels. A diagram of the different signal levels which
should be considered can be seen in Figure 86.




       AWPR – Automatic Wireless Phone Registration                                         xix
                                                                               December 5, 2005




                                  Figure 86. Signal Strength Levels.


Note that the diagram shows five different signal levels. For the purpose of this testing, voice
quality tests should be performed at approximately -50, -60, -70, -80, and -90 dBm.



1.5 Test Results


The main purpose of the voice quality testing that was performed in this project was to quantify
the voice quality of the HP iPAQ 6315 in a wireless VoIP network under various received signal
strength conditions. In addition to the wireless VoIP voice quality tests, GSM voice quality
testing was completed in order to allow a comparison to be made between the iPAQ’s
performance while on the VoIP compared to the GSM network.


First, the voice quality between the PSTN phone and the iPAQ was tested at various signal
levels. Seven voice quality measurements were taken at every received signal strength level and
then the average was calculated. This number of samples was found to be statistically significant
by using the standard deviation from the results and the formula below.




       AWPR – Automatic Wireless Phone Registration                                         xx
                                                                                    December 5, 2005


                                                               2
                                            ⎡z σ ⎤
                                        n = ⎢ a/2 ⎥
                                            ⎣ E ⎦

This formula was found at http://www.isixsigma.com/library/content/c000709.asp. In the

formula, σ is the population standard deviation.          is known as the critical value where the

positive   value is at the vertical boundary for the area of        in the right tail of the standard
normal distribution. A 95% degree confidence interval corresponds to           = 0.05. By using this
value, one can form a normal distribution graph similar to the one seen in Figure 87.




                                Figure 87. Normal Distribution Graph.




In the figure, each of the shaded tails have an area of        = 0.025. The region to the left of
and to the right of   = 0 is 0.5 - 0.025, or 0.475. In the Table of the Standard Normal ( )

Distribution, an area of 0.475 corresponds to a    value of 1.96. The critical value is therefore
= 1.96. This z value corresponds to the error in the tails of a normal distribution. The margin of

error, , is the maximum difference between the observed sample mean             and the true value of
the population mean . In this example




       AWPR – Automatic Wireless Phone Registration                                                 xxi
                                                                                                December 5, 2005


E= ± 0.1. Once all of the necessary values for the formula are known they can be plugged into
the sample formula as seen in the following example.


                                                  2                      2
                                  ⎡z σ ⎤    ⎡ 1.96 ⋅ 0.13 ⎤
                              n = ⎢ a/2 ⎥ = ⎢               = 6.49 = 7 samples
                                  ⎣  E ⎦    ⎣    0.1 ⎥    ⎦

As one can clearly see the when the data from the project is entered into the sample formula, it
was calculated that 6.49 samples would need to be taken. This value can then be rounded up to 7
since a partial sample cannot be taken.


The team began the signal level testing at the -40 dBm level and ended the test at the -90 dBm
level. In between these levels, measurements were taken at every 10 dBm level. The results
from this test can be seen below in Figure 88.



                                     MOS vs. Received Signal Strength
       (Mean Opinion Score)




                              4.50
                              4.00
                              3.50
              MOS




                              3.00                                                              PSTN -> VoIP
                              2.50                                                              VoIP -> PSTN
                              2.00
                              1.50
                              1.00
                                  -100     -90    -80    -70    -60   -50    -40    -30
                                                        RSSI (dBm)


                                         Figure 88. MOS vs. RSSI (PSTN->VoIP and VoIP->PSTN).


As one can clearly see, the MOS scores for the PSTN to VoIP voice quality test remained stable
around 3.4 from the -40 dBm level until the -70 dBm. After reaching the -70 dBm received




      AWPR – Automatic Wireless Phone Registration                                                        xxii
                                                                                 December 5, 2005


signal strength level, the voice quality began to degrade. At -80 dBm, the MOS value dropped to
3.1 and then at -90 dBm the voice quality dropped to 2.9.


Next, the same test was performed for the voice quality with audio going from the iPAQ to the
PSTN. The results can also be seen in Figure 88. In this test, the MOS remained stable
regardless of the received signal strength level. This stability could be due to the fact the voice
quality of the phone was poor even at the best possible signal strength, so it could not degrade
much. One reason that the voice quality was degraded so much for these tests was because of
grounding problems that existed between the iPAQ and the VQT. The grounding problems
caused a minor feedback hum to be present in the audio stream coming from the iPAQ, which
lowered the MOS scores. Although extensive research was performed in an attempt to reduce the
feedback, no resolutions were found for this problem.


Next, the voice quality was tested between the PSTN phone and the GSM and vice versa. These
results, along with the averages from the VoIP tests can be seen in Figure 89.



                                   VoIP vs. GSM Voice Quality

                         4.00
                         3.50
       Average PESQ LQ




                         3.00                                                      PSTN -> VoIP
                         2.50
                                                                                   VoIP -> PSTN
                         2.00
                         1.50                                                      PSTN -> GSM
                         1.00                                                      GSM -> PSTN
                         0.50
                         0.00




                                Figure 89. VoIP vs. GSM Voice Quality.


The final test scenario which was completed was the VoIP to VoIP testing. Similar to the VoIP
to PSTN tests which were performed, these tests results proved to be much lower due to
grounding issues. The results for the VoIP to VoIP testing can be seen in Figure 90.




       AWPR – Automatic Wireless Phone Registration                                           xxiii
                                                                                December 5, 2005




                       MOS vs. Received Signal Strength
(Mean Opinion Score)
                       4.50
                       4.00
                       3.50
       MOS




                       3.00                                                 IPAQ 1 -> IPAQ 2
                       2.50                                                 IPAQ 2 -> IPAQ 1
                       2.00
                       1.50
                       1.00
                           -100 -90 -80 -70 -60 -50 -40 -30
                             Received Signal Strength (dBm)

                                   Figure 90. VoIP to VoIP Voice Quality.




AWPR – Automatic Wireless Phone Registration                                              xxiv
                                                                                December 5, 2005


2. Wired Traffic Loading

2.1 Objective
The purpose of this test was to determine how the voice quality of a phone conversation was
affected when various amounts of traffic were loaded to an access point’s wired interface. The
test began with no traffic load, and the amount of generated traffic was slowly incremented until
an MOS score of 1.0 was seen by the Agilent Voice Quality Tester.

2.2 Equipment


•      Nortel BayStack 460
•      Nortel WLAN 2230
•      Nortel WLAN Security Switch
•      HP iPAQ 6315
•      Agilent Voice Quality Tester
•      J1996 VQT Phone adapter – amplifier to help adjusts the audio input and output for each
       device under test
•      PSTN Phone
•      IXIA Traffic Generator with IxExplorer software


The traffic loading tests were completed using the IXIA traffic generator. The IXIA traffic
generator is a “dumb” traffic generator that was used to impose user-defined traffic to the back of
the access point. The software that controls the traffic generator is called IxExplorer.

2.3 Test Plan


Using the IXIA traffic generator, traffic was loaded to the back at the access point in 150 kbps
increments in order to simulate the amount of traffic for two calls. A single G.711 codec call
utilizes approximately 75 kbps. The actual voice data for a call is approximately 64 kbps while
the overhead added to each call is approximately 11 kbps. The voice quality measurements that
were taken at each rate were performed in two separate directions. First, measurements were



       AWPR – Automatic Wireless Phone Registration                                         xxv
                                                                                 December 5, 2005


taken when audio was sent from the PSTN phone to the iPAQ phone. Then, voice quality
measurements were taken when audio was sent from the iPAQ to the PSTN phone. The RSSI
value received by the handset during the tests was kept at approximate -70 dBm. The test
network for this test should appear similar to the drawing seen in Figure 91.




                              Figure 91. Wired Traffic Loading Network.


As stated previously, the IxExplorer software was used along with the IXIA Chassis to flood the
network access point. In order to begin the test setup, the user should start IxExplorer and wait
for the Chassis light to turn from red to green, indicating that the device is in a ready state. The
introduction screen of IxExplorer should look similar to the screen shown in Figure 92.




       AWPR – Automatic Wireless Phone Registration                                           xxvi
                                                                              December 5, 2005




                             Figure 92. IxExplorer Ready State Change.


Next, the user must login to the program by selecting the Login option from the Multi-user menu
shown in the previous figure. This will bring the user to the login screen shown in Figure 93
below.




                                      Figure 93. Login Screen.




         AWPR – Automatic Wireless Phone Registration                                     xxvii
                                                                               December 5, 2005


Once the user has logged in, he or she must take ownership of the Chassis. This can be done by
right-clicking on the name of the chassis and selecting the Take Ownership option. This action
can be seen in Figure 94.




                              Figure 94. IxExplorer Chassis Ownership.


Next, a new traffic stream should be created, which will be used to generate traffic through the
back of the access point. A new traffic stream can be created by expanding the Chassis tree,
expanding the Port 01 tree, clicking on the Packet Streams option, and then by right-clicking in
the open space on the right side of the screen and selecting New Stream. This process can be
seen in Figure 95.




      AWPR – Automatic Wireless Phone Registration                                         xxviii
                                                                                December 5, 2005




                              Figure 95. IxExplorer New Traffic Stream.


This will bring up the “Stream Properties” window. First, the user should specify the Stream
Control options for the stream. In this screen, the user can specify the name of the stream and
how much traffic should be generated. The percentage of the max rate refers to the amount of
traffic generated in terms of a gigabyte since the port on the chassis is a gigabyte port. The
Stream Properties window can be seen in Figure 96.




       AWPR – Automatic Wireless Phone Registration                                          xxix
                                                                              December 5, 2005




                       Figure 96. Name Traffic Stream and Set Maximum Rate.


Next, the user should click on the “Frame Data” tab to view more details about the traffic stream.
Within this window, the protocols, which should be used at each layer, can be specified. The
user should specify to use an Ethernet II frame on IPv4 and TCP/IP. The preamble and frame
size should be left alone. The completed “Frame Data” tab can be seen in Figure 97.




      AWPR – Automatic Wireless Phone Registration                                         xxx
                                                                               December 5, 2005




                               Figure 97. Traffic Stream's Frame Data.


Before exiting the Stream Properties, the user should click on the Edit button in the Protocols
section to view the window shown in Figure 98. Using this window, the user should specify the
destination address, which should be 165.91.82.71 (the access point), and the source address,
which should be any unused IP address (in this case 165.91.82.81).




      AWPR – Automatic Wireless Phone Registration                                         xxxi
                                                   December 5, 2005




                           Figure 98. IP Header.




AWPR – Automatic Wireless Phone Registration                 xxxii
                                                                              December 5, 2005


After setting the parameters, the user should return to the main window and right click on the
Card 01 and select Open Protocol Window as seen in Figure 99.




                                 Figure 99. Open Protocol Window.




      AWPR – Automatic Wireless Phone Registration                                         xxxiii
                                                                           December 5, 2005


After selecting the Open Protocol Window option, the user will be brought to the window
shown in Figure 100. Within this window, the user should add 165.91.82.1 as the default
gateway address.




                           Figure 100. Adding Default Gateway Address.


Next, the user should right-click in the open space and choose New IP Range as shown in Figure
101.




                                    Figure 101. New IP Range.




       AWPR – Automatic Wireless Phone Registration                                       xxxiv
                                                                               December 5, 2005


The user should then fill in the fields for the new IP range. The local IP address should be
165.91.82.81. This is the same address specified as the source address in the data stream. The
MAC address should be specified as 00 00 00 00 01 0. Then the Use Network checkbox should
be checked, which will allow specification of the network mask width. Since the network mask
is 255.255.255.128 or a \25 network, the Network Mask Width should be set to 25. The
completed Protocol fields can be seen in Figure 102.




                                 Figure 102. Fields for New IP Range.


Next, the user should go to Port 1 in the ARP section of the menu as shown in Figure 103.




      AWPR – Automatic Wireless Phone Registration                                          xxxv
                                                                              December 5, 2005




                                         Figure 103. ARP.


Then, the user should click on the Transmit button. This action will cause the traffic generator
to send an ARP request in order to get the MAC address for the gateway. Once the MAC address
is retrieved, the window shown in Figure 104 should appear.




                                Figure 104. MAC Address Retrieved.




      AWPR – Automatic Wireless Phone Registration                                        xxxvi
                                                                               December 5, 2005


Finally, once that step has been completed, the user should click on Column D in the bottom
window. Column D corresponds with Port 1 of the chassis. Once the selection has been
completed, the user should hit the play button to start generating traffic. This process can be
seen in Figure 105 below.




                                Figure 105. Begin Traffic Generation.




       AWPR – Automatic Wireless Phone Registration                                         xxxvii
                                                                                       December 5, 2005


2.4 Test Results


The voice quality portion of the traffic loading test was completed to determine how the designed
network reacts to saturation. Using the IXIA traffic generator’s IxExplorer software, the traffic
levels were created by altering the percentage of traffic to be sent in relation to maximum rate of
the generator, or the “% Max Rate”. Table 22 displays the proposed generated traffic level
(kbps), actual generated traffic level (Mbps) according to the software, and IXIA “% Max Rate”
setting to create the specific traffic level.


                       Table 22. Generated Traffic Levels on IXIA Traffic Generator.
                                 Generated       Generated
                                                                   IXIA
                                Traffic Level   Traffic Level
                                                                  Setting
                                   (kbps)          (Mbps)

                                     0               0              N/A


                                     75          0.073242188      0.00940


                                    225          0.219726563      0.02798


                                    375          0.366210938      0.04680


                                    525          0.512695313      0.06550


                                    675          0.659179688      0.08430


                                    825          0.805664063      0.10300


                                    975          0.952148438      0.12180


                                    1125        1.098632813       0.14050


                                    1275         1.245117188      0.16000


                                    1425        1.391601563       0.17800


                                    1575         1.538085938      0.19680




       AWPR – Automatic Wireless Phone Registration                                              xxxviii
                                                                                                   December 5, 2005


Seven voice quality tests, or trials, were completed for each traffic level. As previously stated,
the traffic levels were incremented by 150 kbps (75 kbps + 75 kbps), or 2 calls. The only
exception is the first two traffic levels at which the generated traffic was only increased by 75
Kbps. The average of the resulting voice quality levels were calculated and placed into the graph
shown in Figure 106.



                                                          MOS vs. Traffic Load


                                4.25
         (Mean Opinion Score)




                                3.75
                                3.25
                MOS




                                2.75                                                                PSTN -> VoIP
                                2.25                                                                VoIP -> PSTN
                                1.75
                                1.25
                                0.75
                                       0   0.25     0.5    0.75      1      1.25     1.5    1.75
                                              Amount of Traffic Generated (Mbps)


                                                  Figure 106. Voice Quality vs. Traffic Load.


The results of the traffic loading test showed that the quality of the voice from the PSTN phone
to the iPAQ has a semi-consistent quality for traffic loads until about .5Mb of traffic was
generated. At this point, the quality of the voice begins to degrade. At the final recorded rate
(1575 Kbps), the voice quality had degraded far into the unacceptable range with a MOS value of
1.3.


However, the results of the traffic loading test also showed that the quality of the voice from the
iPAQ to the PSTN phone has a semi-consistent quality regardless of the traffic level. The
average MOS values for all of the traffic levels were between 1.0 and 1.88. This is due to the
fact that the generated traffic is being transmitted by the access point to a device on the wireless
network. This does not affect traffic that is being received by the RF interface on the access
point.




         AWPR – Automatic Wireless Phone Registration                                                         xxxix
                                                                                    December 5, 2005


   All of the traffic loading voice quality measurements can be found in Table 23.

                               Table 23. Voice Quality vs. Generated Traffic.
 Number of
                Direction of   PESQ   PESQ     PESQ     PESQ    PESQ     PESQ    PESQ    PESQ          Standard
Simultaneous
                   Test        LQ 1   LQ 2     LQ 3     LQ 4     LQ 5     LQ 6   LQ 7   Average        Deviation
   Calls

               PSTN -> VoIP    3.57    3.29     3.44     3.32    3.63     3.31   3.40    3.42     0.133130799
     1
               VoIP -> PSTN    1.98    2.00     1.89     2.00    1.88     1.89   1.93    1.94     0.053984125
               PSTN -> VoIP    2.76    3.31     3.37     3.35    3.10     3.44   3.54    3.27     0.261005382
     2
               VoIP -> PSTN    1.87    1.95     1.88     1.78    1.51     1.99   1.75    1.82     0.160460646
               PSTN -> VoIP    3.29    3.19     3.48     3.09    3.53     3.42   3.38    3.34     0.158745079
     4
               VoIP -> PSTN    1.83    1.82     1.93     1.95    1.89     1.94   1.80    1.88     0.062716292
               PSTN -> VoIP    3.09    3.49     3.02     3.31    3.64     3.08   3.49    3.30     0.244520912
     6
               VoIP -> PSTN    1.67    1.87     1.94     1.81    1.91     1.69   1.75    1.81     0.106435762
               PSTN -> VoIP    2.76    3.18     3.10     2.83    2.93     2.94   2.98    2.96      0.14525839
     8
               VoIP -> PSTN    1.78    1.70     1.83     2.01    1.80     1.91   1.84    1.84     0.098898698
               PSTN -> VoIP    2.68    2.51     2.35     2.18    2.17     2.52   2.65    2.44     0.208783324
    10
               VoIP -> PSTN    1.52    1.19     1.69     1.32    1.44     1.81   1.22    1.46     0.234297328
               PSTN -> VoIP    1.18    1.27     1.41     1.39    1.14     1.36   1.28    1.29     0.103601802
    12
               VoIP -> PSTN    1.15    1.19     1.55     1.34    1.12     1.64   1.50    1.36     0.210147341
               PSTN -> VoIP    1.00    1.00     1.00     1.01    1.00     1.07   1.00    1.01     0.026095064
    14
               VoIP -> PSTN    1.56    1.15     1.56     1.55    1.43     1.30   1.26    1.40      0.16667619
               PSTN -> VoIP    1.00    1.00     1.00     1.00    1.00     1.00   1.00    1.00             0
    16
               VoIP -> PSTN    1.61    1.23     1.29     1.46    1.19     1.18   1.17    1.30     0.168310937
               PSTN -> VoIP    1.00    1.00     1.00     1.00    1.00     1.00   1.00    1.00             0
    18
               VoIP -> PSTN    1.15    1.12     1.26     1.36    1.00     1.58   1.45    1.27     0.202637373
               PSTN -> VoIP    1.00    1.00     1.00     1.00    1.00     1.00   1.00    1.00             0
    20
               VoIP -> PSTN    1.18    1.12     1.00     1.07    1.00     1.34   1.00    1.10     0.126019651
               PSTN -> VoIP    1.00    1.00     1.00     1.00    1.00     1.00   1.00    1.00             0
    22
               VoIP -> PSTN    1.00    1.00     1.00     1.00    1.00     1.00   1.00    1.00             0




           AWPR – Automatic Wireless Phone Registration                                           xl
                                                                                   December 5, 2005


3. Wireless Traffic Loading

3.1 Objective


The purpose of this test is to load traffic on the wireless interface of an access point and see how
the traffic load affects the voice quality of a call. The test will begin with no traffic load, and the
traffic will be added in small increments in order to determine any differences in voice quality
scores.



3.2 Equipment


•         Nortel BayStack 460
•         Nortel WLAN 2230
•         Nortel WLAN Security Switch
•         HP iPAQ 6315
•         Agilent Voice Quality Tester
•         J1996 VQT Phone adapter – amplifier to help adjusts the audio input and output for each
          device under test
•         PSTN Phone
•         IXIA Traffic Generator with IxExplorer software
•         IXIA IxWLAN
•         Linksys WET11 Wireless Ethernet Bridge




       AWPR – Automatic Wireless Phone Registration                                             xli
                                                                                December 5, 2005



3.3 Test Plan
3.3.1 Wireless Bridge


One of the wireless traffic loading tests that were attempted involved a Linksys WET11 Wireless
Ethernet bridge. The test setup can be seen in Figure 107 below.




                                Figure 107. Wireless Bridge Test Setup.


As one can clearly see, the IXIA traffic generator and the wireless bridge are placed on VLAN 3
while the rest of the network remains on VLAN 6. This is done in order to ensure that the
generated traffic can only reach its host through the wireless portion of the network. If this task
is not performed, traffic can reach the destination host by merely traveling through the wired
portion of the network.


In order to begin the configuration portion of this test, the network bridge has to be configured
first. After connecting to the bridge initially using a PC and a crossover cable, the IP can be
changed to the desired address, 165.91.82.80. Then, after this was completed, the bridge is
accessible to any computer that was placed on VLAN 3. Figure 108 shows a computer
connecting to the bridge via the web interface.




       AWPR – Automatic Wireless Phone Registration                                          xlii
                                                                               December 5, 2005




                                  Figure 108. Bridge's Web Interface.


Once on the bridge’s web interface, the user can easily change any possible options on the
device. On the first setup screen, the user can specify the IP address, subnet mask, and gateway
of the bridge. Once these options have been changed to the desired settings, the user can join the
bridge to an existing wireless network. Figure 109 shows the basic setup screen that the user will
arrive at after logging into the bridge.




                              Figure 109. Basic Configuration of the Bridge.




       AWPR – Automatic Wireless Phone Registration                                          xliii
                                                                               December 5, 2005


As previously stated, once the network settings have been specified for the bridge, the user can
join the bridge to a wireless network. This can be done by pressing the Site Survey button on the
main setup screen. After performing this task the screen shown in Figure 110 will appear.




                                  Figure 110. Site Survey Window.


Once in the Site Survey window, the user should click on the AWPR SSID, which will cause the
window shown in Figure 111 to pop-up.




      AWPR – Automatic Wireless Phone Registration                                          xliv
                                                                              December 5, 2005




                                Figure 111. Site Survey Confirmation.


When this prompt arrives the user should select the OK option and then a message which
indicates that the user will need to enter the WEP key. The WEP key warning box is shown in
Figure 112.




                              Figure 112. WEP Key Security Warning.


Next, the user will arrive at the WEP key screen. The user should select 1 as the Default
Transmit Key. Then, the user should choose to use 64-Bit WEP Encryption. After that, the user
should enter the WEP Key (ABCDEF1234) into the Key 1 input field. The completed WEP key
screen can be seen in Figure 113.




      AWPR – Automatic Wireless Phone Registration                                          xlv
                                                                                December 5, 2005




                                   Figure 113. WEP Key Window.


Once the user has completed the WEP key entry, he or she should click the Apply button,
causing the bridge to reboot. The user should wait approximately one minute and then reconnect
to the web interface. Once the user arrives at the web interface, he or she should select the
Advanced option which is located on the top main menu. This will bring the user to the screen
shown in Figure 114.




       AWPR – Automatic Wireless Phone Registration                                         xlvi
                                                                             December 5, 2005




                               Figure 114. Advanced Bridge Settings.


In the Advanced Settings window, the user should confirm that the Transmission Rate option is
set to Auto, the Authentication Type is set to Open System, and the Cloning Mode is disabled.
The correct option choices are shown in Figure 114. After the user clicks the Apply button, the
bridge should reboot again. The user should wait one minute and then re-renter the web
interface. When the user reaches the main web interface window, he or she should select the
Status option from the main menu. This will bring the user to the screen shown in Figure 115.




      AWPR – Automatic Wireless Phone Registration                                       xlvii
                                                                                December 5, 2005




                                     Figure 115. Status Window.


Once in this screen, the user should make sure that the wireless settings are all correct. The SSID
should be AWPR, the Network Type should be Infrastructure, the Channel should be 9, and WEP
should be enabled. These options are all shown in the previous figure. If everything is correct,
then the bridge setup is complete, and the switch must be setup to allow the bridge to
communicate with the traffic generator.


The bridge works by replicating everything that it sees on its wired port to a wireless access
point. By default, the switch will not know that the bridge is connected to an access point.
Therefore, the switch will not forward anything to the bridge’s wired port. This problem can be
fixed rather easily. First, the user should login to the switch and enter the global configuration
mode. Then the user should enter the commands shown in Figure 116.




                                    Figure 116. Switch Commands.




       AWPR – Automatic Wireless Phone Registration                                          xlviii
                                                                                 December 5, 2005




Although this command is dependent on where the traffic generator and bridge are plugged in, it
is a very simply concept. The first command sets up a monitor session in which the source is Gi
0/2 and the mode is rx, or receive. This means that everything that is received on Gi 0/2 will be
forwarded to wherever the user specifies. Currently Gi 0/2 is the traffic generator’s switch port.
Next, the second command specifies that everything received by the source port should be sent to
the Fa 0/11 interface. This is the port at which the bridge is plugged into. Also note that the
destination is placed in ingress mode, meaning that the port will still be allowed to perform
functions other than monitoring. Once this command is completed, the user should wait a couple
of minutes to ensure that the network is fully up and running.


After a couple of minutes have passed, the user can log into the Nortel Security switch and view
all of the clients that are on the wireless network. The user should look for the MAC address of
the bridge, which can be seen on the main bridge configuration page. Once the bridge is found,
the user should perform a Link Test. After the user clicks the Link Test button next to the correct
MAC address, he or she should wait a couple of seconds until the results appear. This test can be
seen in Figure 117




                                     Figure 117. Bridge Link Test.


If all or almost all of the packets are sent and received, then the link between the bridge and the
access point is operational, and the user can configure the traffic generator. Within the


       AWPR – Automatic Wireless Phone Registration                                          xlix
                                                                                 December 5, 2005


IxExplorer traffic generator software, the user should setup a continuous packet stream to operate
at a 2.625% Max Rate, equivalent to 20 Mbps. This portion of the traffic generating setup can be
seen in Figure 118.




                                Figure 118. IXIA Traffic Stream Setup.


Next, the user should setup the traffic stream to have 64 byte Ethernet II UDP/IP frames with no
errors. This portion of the traffic generator setup can be seen in Figure 119.




       AWPR – Automatic Wireless Phone Registration                                        l
                                                                               December 5, 2005




                                 Figure 119. IXIA Frame Data Setup.




Next, the user should specify the Destination Address of the traffic to be 165.91.82.54, which is a
workstation located on VLAN 3. The Source Address should then be set to 165.91.82.83, a
random address on the same network. This portion of the traffic generator setup can be seen in
Figure 120.




      AWPR – Automatic Wireless Phone Registration                                         li
                                                                                December 5, 2005




                                  Figure 120. IXIA IP Header Setup.


As the final portion of the IXIA traffic generator setup, the user should specify the destination
and source MAC addresses for the traffic stream. The destination MAC address should be setup
to the MAC address of the 165.91.82.54 workstation, 00:0D:56:91:AB:E7. Then the source
MAC address should be setup to any random value. The MAC address setup for the IXIA traffic
generator can be seen in Figure 121.




       AWPR – Automatic Wireless Phone Registration                                          lii
                                                                               December 5, 2005




                                Figure 121. IXIA MAC Address Setup.


After the final step has been performed for setting up the IXIA traffic generator, the user should
ensure that the traffic generator can communicate with the workstation. This can be done by
sending a ping to 165.91.82.54. An example of this ping test can be seen in Figure 122.




                                     Figure 122. IXIA Ping Test.


       AWPR – Automatic Wireless Phone Registration                                         liii
                                                                                   December 5, 2005




If the ping test is successful, then the user can begin the traffic loading test, by pressing the play
button. If traffic is being sent, then the frames sent field should be increasing at a fairly large
rate. An example of the traffic generation stats window can be seen in Figure 123.




                           Figure 123. IXIA Traffic Generator Stats Window.


After the traffic generator is started, the user should repeat the link test from the Nortel Security
Switch. Since the traffic generator is set to generate approximately 20 Mb the link should be
completely flooded and the user should not receive any replies to the pings it sends. An example
of a flooded ping test between the access point and bridge can be seen in Figure 124.




       AWPR – Automatic Wireless Phone Registration                                             liv
                                                                                December 5, 2005




                                    Figure 124. Flooded Ping Test.


The flooding of the links between the access point and the bridge will not affect any of the other
wireless connections. This can be done by performing a link test between any of the other clients
and the access point. If the wireless network is working properly, all or most of the packets
should be received. An example of this test can be seen in Figure 125.




                               Figure 125. Link Test for Another Client.


Next, in order to ensure that the generated packets are being transmitted across the air, a laptop
with a wireless Ethernet card should be used to capture packets in Ethereal. When this
measurement is performed one can clearly see that the bridge is not outputting a total of 20




       AWPR – Automatic Wireless Phone Registration                                          lv
                                                                                 December 5, 2005


Mbps. In fact, as seen in Figure 126 one can see that the packet capture only shows that
approximately 50,000 bytes are being transmitted.




                                 Figure 126 Ethereal Wireless Capture.


This amount of wireless traffic has no effect on the voice quality in either direction.


3.3.2 IxWLAN


The other device that is used in an attempt to load the wireless interface of the access point is the
IXIA IxWLAN device. This device can be used along with the IXIA traffic generator to create
virtual workstations which place traffic on the access points. Like in the previous example, the
IxWLAN and the IXIA traffic generator should be placed on VLAN 3 while the rest of the
network is placed on VLAN 6. A picture of the test network can be seen in Figure 127.




                                  Figure 127. IxWLAN Test Network.


       AWPR – Automatic Wireless Phone Registration                                           lvi
                                                                                December 5, 2005




In order to begin the IxWLAN setup, the user should telnet into the device. Once the user logs
into the device via the telnet session, the user can set the IP address and subnet mask of the
IxWLAN. An example of the telnet session can be seen in Figure 128.




                                 Figure 128. IxWLAN Telnet Session.


After the network settings have been set, the user can login to the web interface of the IxWLAN.
An example of logging into the web interface can be seen in Figure 129.




                                     Figure 129. IxWLAN Login.




       AWPR – Automatic Wireless Phone Registration                                          lvii
                                                                                December 5, 2005




After the user logs into the device, the startup wizard will appear as seen in Figure 130.




                                 Figure 130. IxWLAN Startup Wizard


After the user reaches the IxWLAN wizard, he or she should select the New Scenario option in
order to reach the screen shown in Figure 131.




       AWPR – Automatic Wireless Phone Registration                                          lviii
                                                                                 December 5, 2005




                                       Figure 131. vSTA Screen.


Once the user arrives at this screen, he or she will be able to create one or more virtual station
groups. Since an external source will be used for the traffic generation, it is only necessary to
create a single group containing one virtual station. The user can do this by clicking on the New
Group icon, located on the left menu. After the user clicks the button, the screen in Figure 132
will appear.




       AWPR – Automatic Wireless Phone Registration                                           lix
                                                                               December 5, 2005




                      Figure 132. New IxWLAN Group Window (vSTA Settings).


Within this window, the user should specify to create one virtual station. The user should set the
starting IP address of the station to be 165.91.82.82 and the starting MAC Address of the station
to be 00:0D:56:83:00:00. Next, the user should click on the Traffic tab in order to reach the
screen shown in Figure 133.




      AWPR – Automatic Wireless Phone Registration                                         lx
                                                                                December 5, 2005




                      Figure 133. New IxWLAN Group Window (Traffic Settings).


In the Traffic settings, the user should set the Traffic Source field to External and Layer 2. Then,
the user should click on the Security tab in order to reach the window shown in Figure 134.




       AWPR – Automatic Wireless Phone Registration                                         lxi
                                                                                December 5, 2005




                     Figure 134. New IxWLAN Group Window (Security Settings).


In the Security window the user should set the Encryption option to On and then they should
enter the WEP key (abcdef1234) as shown in the previous figure. The user should then click the
Create button to complete the Virtual Station setup. After that step is completed, the user should
click the IxWLAN tab on the left side of the screen. Once this tab expands, the user should click
the Join SUT. This button is shown in Figure 135.




      AWPR – Automatic Wireless Phone Registration                                         lxii
                                                                                December 5, 2005




                                        Figure 135. Join SUT.


This will allow the user to associate the IxWLAN with one of the project network access points.
Once the button is clicked, the user will have to select the AWPR network and then re-renter the
WEP key. Then the IxWLAN will join the wireless network. After this step is completed, the
web interface will return to the main screen, and the user can confirm that the IxWLAN has
joined the SUT (System Under Test) by observing the status indicator in the upper right hand
portion of the screen. Once the user has confirmed that the IxWLAN is on the network, the user
should start running the virtual station, by clicking on the play button. After the station has been
started, the user should see a screen similar to the one shown in Figure 136.




                               Figure 136. Running the Virtual Station.


Next, before anything else can be done on the IxWLAN, the IXIA traffic generator should be
setup. As shown in the previous wireless test, the test stream should be set to 20 Mbps.
However, in this test the Source IP Address should be set to 165.91.82.82, which is the address of
the virtual station. The Destination IP Address should remain at 165.91.82.54. This step can be
seen in Figure 137.




       AWPR – Automatic Wireless Phone Registration                                          lxiii
                                                                            December 5, 2005




                                Figure 137. IXIA IP Address Setup.


After the IP address has been changed, the user should change the Source MAC Address to the
same address that was given to the virtual station. The Destination MAC Address should remain
the MAC address for the given destination IP Address. This screen can be seen in Figure 138.




      AWPR – Automatic Wireless Phone Registration                                      lxiv
                                                                                 December 5, 2005




                                Figure 138. IXIA MAC Address Setup.


After the traffic generator setup has been completed, the user should attempt a ping from the
traffic generator in order to ensure connectivity across the wireless link. In addition, the user
should attempt the link test from the Nortel Security Switch to ensure that the link between the
access point and the IxWLAN is working properly. Then, the user should begin generating
traffic. Once the traffic has been started, the user should return to the IxWLAN web interface
and click on the Monitors tab on the left menu and select the New Monitor option. This option
can be seen in Figure 139.




                                      Figure 139. New Monitor.




       AWPR – Automatic Wireless Phone Registration                                           lxv
                                                                                  December 5, 2005


This will bring the user to a New Monitor Window, where he or she will be allowed to pick
which fields that they would like to view. For the purpose of this test, the user should select to
monitor the “Total Received Data Frames” and “Total Transmitted Data Frames” fields. After
these options are selected, the user should click the Create button. The New Monitor setup
window can be seen in Figure 140.




                                   Figure 140. New Monitor Window.


After the monitor is created, it will appear at the bottom of the screen, and it will update in real-
time. The total numbers of packets received and transmitted correspond with wireless packets
only, so when the external traffic generator is running, the user should see the transmitted packets
counter increment fairly quickly. An example of the monitors can be seen in Figure 141.




       AWPR – Automatic Wireless Phone Registration                                            lxvi
                                                                                December 5, 2005




                                    Figure 141. Monitors Display.


In order to ensure that the packets are being transmitted wirelessly, the user can capture packets
using a wireless Ethereal card and Ethereal. After capturing the wireless packets that were
transmitted when 20 Mbps was generated with the IXIA traffic generator, approximately 100,000
bytes per second was seen coming from the IxWLAN. Although this is more data than what was
being generated wirelessly by the wireless bridge, this amount still has not effect on the voice
quality of the call. A graph of the Ethereal capture that was seen by the laptop can be seen in
Figure 142.




                               Figure 142. IxWLAN Ethereal Capture.




       AWPR – Automatic Wireless Phone Registration                                         lxvii
                                                                               December 5, 2005


3.4 Test Results


After loading the network with the maximum of amount of traffic from the wireless bridge and
from the IxWLAN device, there was no change in the voice quality of a call in either direction.
This is different than what was seen in the wired traffic test. Apparently there is some form of
wireless client throughput limitation on the Nortel Security Switch. The project team contacted a
Nortel representative but was not able to determine the direct cause of the problem.




      AWPR – Automatic Wireless Phone Registration                                         lxviii
                                                                                 December 5, 2005



D. Battery Life

1. Objective



The objective of this test is to identify the life of the HP iPAQ 6315 battery. The battery life will
be recorded in terms of hours. The battery life test includes five individual tests. First, the
battery will be tested in talk time and idle mode while the phone is on the cellular GSM network
with the 802.11 radio turned on. Next, the idle and talk time will be recorded while on the
cellular GSM network with the 082.11 radio turned off. This will give a good comparison of
how much battery is wasted by this radio on the GSM network. Finally, the battery life will be
recorded while the iPAQ is in talk mode and idle mode on an 802.11b network.


2. Equipment
Hardware
       Laptop running Windows XP
       HP iPAQ 6315
       Logitech Webcam


Software
       Camtasia Studio




       AWPR – Automatic Wireless Phone Registration                                           lxix
                                                                                          December 5, 2005


3 GSM Testing Procedure

3.1 GSM Idle Mode

(3.1.1) First, the user should fully charge the HP iPAQ battery following the manufacturer’s charging
instructions.
(3.1.2) The iPAQ is fully charged when the battery indicator shows 100%.
(3.1.3) If the iPAQ is turned off, the power indicator light at the top of the iPAQ will be solid orange to
indicate that the battery is fully charged.
(3.1.4) The user should place the iPAQ in the cradle while it is plugged into the wall.
(3.1.5) Next, the user should ensure that the 802.11 radio is turned on.
(3.1.6) The user should setup a webcam so that it will record the iPAQ screen. Make sure that the time
and mode can be seen in the recording.
(3.1.7) Next, the user should start the Camtasia Studio software and click to start a new project by
recording the screen. Figure 143 shows this step.




                                 Figure 143. Start a New Project in Camtasia.


Note: If Camtasia Studio is not available, any software to record from a webcam can be used.


(3.1.8) The user should select to record the desired area of the screen.
(3.1.9) All of the default settings for all of the other recording options should be used until reaching the
screen shown in Figure 144. On this screen, select the Record Camera and Camera Preview option, and
then hit Next.




       AWPR – Automatic Wireless Phone Registration                                                    lxx
                                                                                    December 5, 2005




                                 Figure 144. Camtasia Recording Options.


(3.1.10) The user should hit Next one more time and click Finish.
(3.1.11) Next, the user should go to the Tools menu and select Options under the Streams tab, the user
should change the capture frame rate to 1 frame/second, as shown in Figure 145. This will dramatically
decrease the size of the recorded video.




                                 Figure 145. Camtasia Recording Options.




       AWPR – Automatic Wireless Phone Registration                                              lxxi
                                                                                    December 5, 2005




(3.1.12) The user should then select the Video Setup option and change the compression to DivX 6.0
Codec.
(3.1.13) The user should click OK twice.
(3.1.14) As a final step, the user should remove the power plug from the iPAQ cradle, while
simultaneously clicking the record button.




                                  Figure 146. Camtasia Recording Mode.


The recording software, shown above in Figure 146, will record the iPAQ screen as the battery
drains. This will allow the team to ensure that the phone stayed in GSM mode the entire time.
Once the iPAQ’s battery has been completely depleted, the user should press the Video Stop icon
to stop the recording. The time displayed on the iPAQ at the beginning and end of the test video
will indicate the battery life of the phone in GSM mode. The user should enter these values into
the spreadsheet, which will be used to estimate the battery life of a caller based on the caller’s
percentage of talk time per hour.




         AWPR – Automatic Wireless Phone Registration                                           lxxii
                                                                                       December 5, 2005


3.2 GSM Dedicated Mode

(3.2.1) The user should complete steps 3.1.1 through 3.1.12.
(3.2.2) Then, the user should initiate a GSM call.
(3.2.3) Next, the user should place a speaker or headphone up to the iPAQ microphone and repeatedly
play an audio file.
(3.2.4) Finally, the user should complete the remainder of the steps in section 3.1.



3.3 GSM Idle Mode (with 802.11 radio off)


(3.3.1) Complete steps 3.1.1and 3.1.2.
(3.3.2) Complete the remainder of the steps in section 3.1, omitting 3.1.3.



3.4 GSM Dedicated Mode (with 802.11 Radio Off)

(3.4.1) Complete steps 3.1.1 through 3.1.2.
(3.4.2) Complete steps 3.1.4 through 3.1.12.
(3.4.3) Initiate a GSM call.
(3.4.4) Place a speaker or headphone up to the iPAQ microphone and repeatedly play an audio file.
(3.4.5) Complete the remainder of the steps in section 3.1.



4. Testing Procedure for 802.11B


The 802.11B parameters will be configured and controlled by Ringer Communications. For all
VoIP calls, the team will be using the G.711 codec, which utilizes a 64 Kbps per call. The
received signal level, RSSI, must be at least -75dBm for accurate testing. All VoIP battery life
testing will be performed on channel 9, which operates at 2.452 GHz.




       AWPR – Automatic Wireless Phone Registration                                              lxxiii
                                                                                   December 5, 2005




4.1 VoIP Idle Mode
(4.1) Repeat all steps in 3.1 while connected to the VoIP network rather than the GSM network.

4.2 VoIP Dedicated Mode
(4.2.1) Complete steps 3.1.1 through 3.1.12 for a VoIP network.
(4.2.2) Initiate a VoIP call.
(4.2.3) Place a speaker or headphone up to the iPAQ microphone and repeatedly play an audio file.
(4.2.4) Complete the remainder of the steps in section 3.1.



5 GSM Test Results

5.1 GSM Idle Mode


Three different tests were performed to determine the GSM standby time. The three recorded
GSM idle mode values from the tests were 4 hours 18 minutes, 4 hours 12 minutes, and 4 hours
15 minutes. These values, as well as the ones for the other tests, can be seen in the table located
in section 6.0. Originally, the team predicted that the phone would last several days while in
GSM standby mode. This prediction was made based on three assumptions. First, GSM has a
sleep mode, which allows the phone to periodically transmit and receive messages from the
network instead of continuously transmitting when a call is not in progress. Second, GSM
operates at a much lower frequency than 802.11 and therefore it consumes considerably less
power. Third, it was assumed that the 802.11 radio would be turned off or placed in a standby
state when the phone entered GSM mode. The results from this test can be seen in Table 24.


                                     Table 24. GSM Idle Mode Times.
                                            Measured Times
                                                   4:18
                                                   4:12
                                                   4:15




       AWPR – Automatic Wireless Phone Registration                                              lxxiv
                                                                                     December 5, 2005


The VeriSign software client performed different than the team originally anticipated. In
retrospect, in order for Ringer Communications to achieve its goal of an automatic registration
process, it is more logical to leave the 802.11 radio turned on while the phone is in GSM mode.
By doing this, the phone will be ready to automatically switch over to a VoIP network if one
becomes available. This will prevent a user from having to worry about whether or not they are
in an 802.11 wireless service area. Instead, the phone will automatically attempt to connect to a
wireless network and register the phone, if a wireless network is available.

5.2 GSM Dedicated Mode


While in dedicated GSM mode, or talk mode, both radios will remain on like they did in GSM
idle mode. The dedicated GSM mode test was performed twice in order to find a good
estimation of the battery life while in GSM mode. The results from the test were times of 3 hours
3 minutes, and 3 hours 4 minutes. This time is significantly shorter than the GSM standby.
Because both the GSM and 802.11 radios were fully operational for the entire duration of this
test, the battery life results for this test were the shortest out of all of the tests. The results from
this test can be viewed in Table 25.


                                  Table 25. GSM Dedicated Mode Times.
                                            Measured Times
                                                  3:03
                                                  3:04



5.3 GSM Idle Mode (with 802.11 radio off)


This test was done in GSM mode with the 802.11 radio turned off. This test was performed
twice to give a good estimation of the battery life without the 802.11 radio on. The results from
this test were 17 hours, 19 minutes and 17 hours, 7 minutes. This test was longer than the other
two because GSM Idle mode has a stand by function that works to save battery when not in use.
This test actually lasted about 13 hours longer than the same test with the 802.11 radio turned on.
This shows how much battery is wasted by the 802.11 radio. From this, it can be assumed that to



       AWPR – Automatic Wireless Phone Registration                                               lxxv
                                                                                 December 5, 2005


enhance battery life, some sort of idle mode should be implemented for the 802.11 mode, like
seen in the GSM mode. The results from this test can be seen in Table 26.


                            Table 26. GSM Idle Mode (with 802.11 radio off).
                                          Measured Times
                                                17:19
                                                17:07



5.4 GSM Dedicated Mode (with 802.11 radio off)


This test, like the previous, was performed in GSM mode with the 802.11 radio turned off. It
was also performed twice to give a good estimation of the battery life talk time, without the
802.11 radio on. The results from this test were 9 hours, 2 minutes and 8 hours, 58 minutes.
This test was slightly longer than the other two because GSM Idle mode has a sleep mode
function that works to save battery when not in use. The results from this test can be seen in
Table 27.


                         Table 27. GSM Dedicated Mode (with 802.11 radio off).
                                          Measured Times
                                                 9:02
                                                 8:58



6 802.11B Test Results

6.1 VoIP Idle Mode


First, the phone was tested in wireless VoIP idle mode. During this test the 802.11b radio
remained on while the GSM radio was automatically turned off by the VeriSign client. The
times recorded for this section were very similar to the times seen in the GSM standby time.
This is due to the fact that GSM uses very little battery life in standby due to its sleep mode.




       AWPR – Automatic Wireless Phone Registration                                           lxxvi
                                                                                   December 5, 2005


Three separate tests were performed while the iPAQ was in idle mode with a received signal
strength level of -40 dBm. In these tests the phone battery lasted 4 hours 15 minutes, 4 hours 20
minutes, and 4 hours 18 minutes. Next, the second set of tests were performed at a received
signal strength of -70 dBm in order to determine the dependence of the iPAQ’s battery on
received signal strength. During this test, the iPAQ lasted 4 hours 20 minutes and 4 hours 18
minutes. This proves that the received signal level is not a key contributing factor to the battery
life. The results from this test can be seen in Table 28.


                                       Table 28. VoIP Idle Mode.
                                   Signal Level      Measured Time
                                        -40                 4:15
                                        -40                 4:20
                                        -40                 4:18
                                        -70                 4:18
                                        -70                 4:20



6.2 VoIP Dedicated Mode


This final test performed was the dedicated mode VoIP test. The results showed 4 hours 1
minute and 3 hours 55 minutes. This talk time is longer than the GSM talk time test due to the
fact that only the 802.11 radio is fully functional for the entire duration of this test. The test
results, however, show that the battery does not last as long in the dedicated VoIP mode. This is
due to the continuous transmission of packets across the wireless network. When the phone is in
idle VoIP mode, it is only sending and receiving update broadcast messages to the network and
to the access point. Similar to the idle mode VoIP test, the GSM radio remains off while the
phone is in dedicated mode on the VoIP network. The results from this test can be seen Table
29.




       AWPR – Automatic Wireless Phone Registration                                             lxxvii
                                                                                       December 5, 2005


                                     Table 29. VoIP Dedicated Mode.

                                           Measured Time
                                                  3:55
                                                  4:01


7 Duty Cycle


The average test results from all of these tests can be seen in Table 30.


                                     Table 30. Average Test Results.
                                                    VoIP           VoIP (with
                                        GSM
                                                  (normal)      802.11 radio off)
                         Idle              4:15       4:18           17:13
                         Dedicated         3:03       3:58             9:00



Next, the duty cycles for each device and scenario can be calculated. The duty cycle represents
different levels of usage for a user. The duty cycle calculation results can be seen in Table 31.


                                         Table 31. Duty Cycles.
                       Talk Time      GSM Duty      VoIP Duty       VoIP (with radio
                      (Dedicated)       Cycle         Cycle         off) Duty Cycle
                          0%             4:15            4:18            17:13
                         10%             4:07            4:16            16:23
                         20%             4:00            4:14            15:34
                         30%             3:53            4:12            14:45
                         40%             3:46            4:10            13:55
                         50%             3:39            4:08            13:06
                         60%             3:31            4:06            12:17
                         70%             3:24            4:04            11:27
                         80%             3:17            4:02            10:38
                         90%             3:10            4:00             9:49
                         100%            3:03            3:58             9:00




       AWPR – Automatic Wireless Phone Registration                                              lxxviii

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:12
posted:2/9/2012
language:
pages:198