Docstoc

Introduction to Wireless LAN _WLAN_

Document Sample
Introduction to Wireless LAN _WLAN_ Powered By Docstoc
					                                                        Wireless LAN


Introduction to Wireless LAN (WLAN)
Wireless technology has been available for a long time and it is only until recent few
years that wireless LAN (WLAN) becomes popular. WLAN has continued to grow at
an incredible rate. Due to the convenience, availability, and cost of wireless hardware,
there is an explosive growth in WLAN deployment and manufacture of WLAN
hardware. It is therefore necessary to have organizations such as FCC, IEEE, the
Wi-Fi Alliance and WLANA to remove barriers of operations between solutions.




Applications of Wireless LANs
Wireless LAN was first used by defense force, then the large enterprises, and now
available to home users. One of the many advantages of WLAN is that it offers a
variety of usages. We will discuss some of the most common uses of WLAN.


a.    Access Role
      Wireless LANs are used as an entry point into wired networks and are mostly
      deployed as an access layer role. It is another method for users to access the
      network. In an OSI reference model, WLAN is a Data Link Layer network. Due
      to lack of speed and resiliency, WLAN is not implemented at the Distribution or
      Core role in a network. The diagram below shows a typical access role of a
      wireless LAN.




D-Link Australia & NZ                                                      Page 1 of 85
Version 1.0
                                                                            Wireless LAN




          Wired                      Access Point
          Network

                                                            Client




                                                             Client

                                        Client

                                                                 Coverage
                                                                 Area


      Figure 1. Access Role of a Wireless LAN


      WLAN offers mobility and is accessible almost anywhere. WLAN solves the
      problem of the need for data cabling and offers users a fast and inexpensive
      solution to stay connected with ability to roam.


b.    Network Extension
      Wireless network can serve as an extension to wired networks. In a typical
      network extension, installation of additional cabling is required and this can cost
      a lot.


      WLAN can be implemented easily to provide seamless connectivity to remote
      areas within a building. As only minimum wiring is required to install a WLAN,
      the costs of hiring installers and purchasing of Ethernet cable is kept to the
      minimal. The diagram below shows an extension of network using WLAN from
      the server farm to the warehouse.




D-Link Australia & NZ                                                       Page 2 of 85
Version 1.0
                                                                         Wireless LAN



                Server Farm




                                                                   Warehouse




      Figure 2. Network Extension


c.    Building-to-Building Connectivity
      In an environment with two or more adjacent buildings, there may be a need to
      have network users in each of these buildings who are accessing to the same
      computer network. A typical way to achieve this is by running cables
      underground from one building to another. Another alternative is by renting
      expensive leased-lines from the local telephone companies.


      By using wireless technology, equipment can be installed quickly and easily to
      two or more buildings in the same network. It can be done without renting
      expensive leased line and digging the ground between buildings. With the use
      of proper WLAN antennas, two or more buildings can be linked together on the
      same network. Although there are some limitations using WLAN, the flexibility
      and cost-saving attracts the network administrators to make use of WLAN.


      There are two different types of building-to-building connectivity. The first is
      called point-to-point (PTP). PTP connection uses a directional antenna at each
      of the end of the link.




      Figure 3. Point-to-Point Connectivity

D-Link Australia & NZ                                                     Page 3 of 85
Version 1.0
                                                                            Wireless LAN




      The other building-to-building connectivity is point-to-multipoint (PTMP). PTMP
      links wireless connections between three or more buildings in a star topology.
      One building serves as the central point of the network and an omni-directional
      antenna is used. The remote buildings will link to the central point where the
      Internet connectivity and server farms are located. Directional antennas are
      used in these buildings.

                                          Omni-
                                        Directional
                                         Antenna                     Directional
          Directional                                                Antenna
          Antenna




                                        Central                         Internet




      Figure 4. Point-to-Multipoint Connectivity


d.    Last Mile Data Delivery
      Wireless Internet Service Providers (WISPs) offer the last mile data delivery
      service to their customers using wireless LAN. ‘Last mile’ refers to the wired or
      wireless communication infrastructure that exists between the central office of
      the telecommunication or cable company and the end user.




                        WISP Tower
                                                             Remote
                                                            Residence

      Figure 5. Last Mile Service

D-Link Australia & NZ                                                        Page 4 of 85
Version 1.0
                                                                           Wireless LAN




      The last mile service is useful at locations where the cable and
      telecommunication companies have difficulties expanding their network to offer
      broadband connections. For example, in rural area, there is no access to the
      broadband connection; it will be more cost effective for the WISP to provide
      wireless access to these areas.


e.    Mobility
      A wireless LAN cannot replace wired LAN in terms of data rates, potential
      WLAN’s intermittent connections and may have higher error rates. Therefore,
      applications designed for a wired LAN may not be suitable in the wireless
      environment. However, WLANs offer an increase in mobility as a trade off for
      speed and quality of service. For instance, staff can perform regular stocktake
      with a wireless hand-held device at any location in the supermarket.

                             Wired Network                   Supermarket




      Figure 6. Mobility of Wireless LAN


      Wireless LAN allows data transfer without requiring time and manpower to
      input data at a wired terminal. Wireless connectivity eliminates the need for
      user devices to be connected using wires.


f.    Small Office, Home Office (SOHO)
      Many users and IT professionals today have more than one computer at home.
      These computers are normally networked together in order to share files,
      printer and Internet access. This configuration is also common in many small
      offices with few employees sharing resources to work more efficiently with
      higher productivity.



D-Link Australia & NZ                                                        Page 5 of 85
Version 1.0
                                                                           Wireless LAN


      Instead of running cables throughout the office or home to create a wired LAN,
      the wireless LAN can provide a simple and effective solution to these small
      offices and homes which are not usually installed with Ethernet ports. Wired
      LAN can create unsightly trunks and holes on the walls and ceiling due to the
      cabling. With wireless LAN, users can be interconnected easily and neatly.


                                                               Print Sharing
              Internet




                                                            Wireless
            Internet         Wireless                       Print Server
            Sharing          Gateway




                                                                      File Sharing


      Figure 7. SoHo Wireless LAN




Wireless LAN Standards
Wireless LAN transmits using radio frequency and it is regulated by the government
bodies. In the United States, the Federal Communications Commission (FCC)
regulates the use of wireless LAN devices. In the current wireless LAN market, there
are several accepted operational standards with are created and maintained by the
Institute of Electrical and Electronic Engineers (IEEE).


These standards are created by a group of specialists that represent many
organizations such as academics, business, military and government. The standards
normally take years to create and agreed upon. These are the latest wireless LAN
standards that are built.


a.    IEEE 802.11
      This is the original wireless LAN standard with the slowest data transfer rate in
      both RF and light-based transmission technologies.


D-Link Australia & NZ                                                      Page 6 of 85
Version 1.0
                                                                            Wireless LAN




b.    IEEE 802.11b
      This standard satisfies a faster data transfer rate and it is a more restrictive
      scope of transmission technologies. It uses the 2.4 GHz frequency bands. This
      standard is also widely promoted as Wi-Fi (certification) by the Wi-Fi Alliance. It
      is an amendment from the original 802.11 standard.


c.    802.11b+
      Texas Instruments (TI) developed a modulation technique called Packet Binary
      Convolution Code (PBCC) that can provide signaling rates of 22 and 33 Mbps.
      TI produces 802.11b-based chipsets that also support 22 Mbps PBCC. They
      are fully 802.11b compliant and when communicating with each other of the
      same standard, it can automatically use the 22 Mbps signaling rate. Another TI
      enhancement is 4X mode, which uses a larger maximum packet size – 4000
      bytes – to reduce overhead and increase throughput.


d.    IEEE 802.11a
      This is a much faster data transfer rate compared to IEEE 802.11b. However, it
      lacks backward compatibility with IEEE 802.11b. It uses the 5 GHz frequency
      bands. It is also an amendment from the original 802.11 standard.


e.    IEEE 802.11g
      This is the most recent standard based on the original 802.11 standard. The
      data transfer rates is as fast as IEEE 802.11a and it is backwards compatible
      with 802.11b.


f.    802.11g+
      This is a Texas Instruments (TI) implementation of the IEEE standard, with
      addition of several vendor-specific capabilities. The 802.11g+ products
      interoperate fully with 802.11b and 802.11g devices.


g.    IEEE 802.11e
      This standard is developed to support Quality of Service (QoS). It will improve
      the capability and efficiency for applications such as voice, video and audio
      transport over the wireless network.




D-Link Australia & NZ                                                       Page 7 of 85
Version 1.0
                                                                         Wireless LAN


Emerging Wireless Standards
There are other emerging standards which have not been released.


a.    IEEE 802.11h
      This standard adds indoor and outdoor channel selection for the 5 GHz bands
      in Europe. This will enhance the channel energy measurement and there is
      reporting mechanisms to reduce interference using Dynamic Frequency
      Selection (DFS) and improve power management using Transmit Power
      Control (TPC).


b.    IEEE 802.11i
      This standard is developed to provide WLAN network security. It uses the
      802.1x port-based EAP standards for user and device authentication and
      Temporal Key Integrity Protocol (TKIP) encryption as the protocol and
      algorithm to improve security of keys used with WEP. It also has two new
      features, pre-authentication and encryption protocols based on Advanced
      Encryption Standard (AES) encryption algorithm. WPA certification, initiated
      from Wi-Fi Alliance, is a subset of 802.11i dealing only with 802.1x/EAP and
      TKIP.


c.    IEEE 802.11k
      This standard defines radio resource measurements for WLAN. Both the
      terminal and access point can make request for information from their peers
      and make decisions about their status and the desired action to be taken. The
      upper layers can also use the measurements to make appropriate decisions.


d.    IEEE 802.11n
      This standard defines standardized modifications to both 802.11 physical layers
      (PHY) and the 802.11 Medium Access Control Layer (MAC) so that modes of
      operation is capable of much higher throughput of at least 100 Mbps
      throughput for applications. It is measured at the MAC data service access
      point (SAP). Other goals are to achieve the throughput with sacrificing range
      and to maintain interoperability with 802.11a and/or 802.11g devices.


e.    WiMAX
      The Worldwide Interoperability for Microwave Access (WiMAX) Forum is a
      coalition of wireless broadband access (WBA) equipment vendors and service

D-Link Australia & NZ                                                     Page 8 of 85
Version 1.0
                                                                           Wireless LAN


      providers to promote the development, refinement and market acceptance of
      IEEE 802.16, a series of standards for fixed wireless broadband (FWB)
      metropolitan area networks (MANs) operating at frequency from 10 to 66 GHz.


f.    IEEE 802.11r
      This standard is targeted to minimize the terminal transfer from one access
      point to another. The goal is to achieve a fast BSS (Basic Service Set)
      transition time that is compatible with applications such as VoIP. It has to
      accomplish the goal without reducing the security features and affecting the
      existing station services.




Radio Frequency (RF) Fundamentals
Radio frequencies are high frequency alternating current (AC) signals that pass along
a copper conductor and then radiated into the air via an antenna. An antenna
transforms the wired signal to a wireless signal and vice versa. When the high
frequency AC signal is radiated into the air, it forms radio waves. These radio waves
propagate from the antenna in a straight line in all directions at once.

                             Radio Waves




                                                         Wireless
                                                         Signals




                                                          Wired
           AC Signals                                     Signals




Figure 8. Radio Frequency


Units of Measurement
There are a few standard units of measurement that a wireless network administrator
need to be familiar with in order to be effectively implementing and troubleshooting
wireless LANs.


D-Link Australia & NZ                                                      Page 9 of 85
Version 1.0
                                                                           Wireless LAN




a.    Watts (W)
      The basic unit of power is a watt. A watt is defied as one ampere (A) of current
      at one volt (V). One watt is equal to an Ampere multiplied times a Volt.


b.    Milliwatt (mW)
      When implementing wireless LANs, the power levels as low as 1 milliwatt
      (1/1000 watt) can be used for a small area and the power levels on a single
      wireless LAN segment are rarely above 100mW. This is enough to
      communicate up to 800 meters in optimum conditions.


      Access points generally have the ability to radiate 30 to 100 mW of power,
      depending on the manufacturer. In some cases of point-to-point outdoor
      connections between buildings, the power levels may reach above 100 mW.
      Most of the power levels referred to will be in mW or dBm. These two units of
      measurement both represent an absolute amount of power and are both
      industry standard measurements.


c.    Decibels (dB)
      When a receiver is very sensitive to RF signals, it may be able to pick up
      signals as small as 0.000001 mW. This number is too small and is likely to be
      ignored or misread. Decibels represent these numbers by making them more
      manageable and understandable. Decibels are based on a logarithmic
      relationship to watts. Concerning RF, a logarithm is the exponent to which the
      number 10 must be raised to reach some given value.


      For example, log 1000 = 3 because 103 = 1000, where 3 is the exponent. The
      logarithm of a negative number or of zero does not exist. Decibels are a relative
      measurement unit unlike the absolute measurement of milliwatts.


d.    dBm
      The reference point that relates the logarithmic dB scale to the linear watt scale
      is: 1 mW = 0 dBm


      The m in dBm refers to the reference of 1 mW, which means a dBm is a
      measurement of absolute power.



D-Link Australia & NZ                                                     Page 10 of 85
Version 1.0
                                                                              Wireless LAN


      The relationship between the decibels scale and watt scale can be estimated
      using the following rules of thumb:




      +3 dB will double the watt value:
                  (10mW + 3dB ≈ 20 mW)


      -3 dB will halve the watt value:
                  (100mW – 3dB ≈ 50 mW)


      +10 dB will increase the watt value by ten-fold:
                  (10 mW + 10dB ≈ 100mW)


      -10 dB will decrease the watt value to one-tenth of that value:
                  (300 mW – 10dB ≈ 30 mW)


      These rules will allow quick calculation of milliwatt power levels when given
      power levels, gains and losses in dBm and dB. The following chart shows the
      reference point is always the same, but the power levels can move in either
      direction from the reference point depending on whether they represent a
      power gain or loss.


            -40       -30    -20     -10     0       +10     +20        +30      +40
           dBm       dBm    dBm     dBm     dBm      dBm     dBm        dBm      dBm




          100         1     10      100      1      10       100    1,000      10,000
          nW         uW     uW      uW      mW      mW       mW      mW         mW



           -12       -9      -6      -3      0       +3      +6          +9     +12
          dBm       dBm     dBm     dBm     dBm     dBm     dBm         dBm     dBm




          62.5       125    250     500      1       2        4          8       16
          uW         uW     uW      uW      mW      mW       mW         mW       mW



      Figure 9. Power Level Chart



D-Link Australia & NZ                                                         Page 11 of 85
Version 1.0
                                                                            Wireless LAN


e.    dBi
      Gain and loss are measured in decibels which quantify the gain of an antenna.
      The decibel unit is represented by dBi. The unit of measurement, dBi, refers
      only to the gain of an antenna. The “i" stands for “isotropic”, which means that
      the change in power is referenced against an isotropic radiator. An isotropic
      radiator is a theoretical ideal transmitter that produces useful electromagnetic
      field output in all directions with equal intensity, and at 100-percent efficiency,
      in three-dimensional space.




Radio Frequency Behaviours
Radio frequency acts inconsistently under given circumstances. Things such as
connector not properly tightened or slight impedance mismatch on the line can cause
erratic behavior and undesired results. The following are some types of behaviors
that can happen to radio waves when they are transmitted.


a.    Gain
      Gain is the term that describe an increase in an RF signal’s amplitude. Gain is
      normally an active process. An external power source such as a RF amplifier is
      used to amplify the signal. Alternatively, a high-gain antenna is used to focus
      the beamwidth of the signal and hence it will increase its signal amplitude.

               Gain as seen by an                       Gain of DSSS as seen by a
                  oscilloscope                              spectrum analyzer
                                Peak Amplitude after Gain

                               Peak Amplitude before Gain




      Figure 10. Power Gain
      For passive processes, it can also cause gain. For example, reflected RF
      signals can combine with the main signal to increase the main signal’s strength.
      Increasing the RF signal’s strength may have positive or negative result.
      Typically, more power is better. However, when a transmitter is radiating power
      very close to the legal power output limit, adding more power will cause a
      serious problem.


D-Link Australia & NZ                                                       Page 12 of 85
Version 1.0
                                                                               Wireless LAN




b.    Loss
      Loss is described as decrease in an RF signal’s amplitude. Many things can
      cause RF signal loss. Resistance of cables and connectors causes loss when
      converting AC signal to heat. Impedance mismatches in the cables and
      connectors can cause power to reflect back towards the source and result in
      signal degradation. An object that is blocking the propagated wave’s
      transmission path can be absorbed, reflected or destroyed.

               Gain as seen by an                         Gain of DSSS as seen by a
                  oscilloscope                                spectrum analyzer
                                  Peak Amplitude before Gain

                                  Peak Amplitude after Gain




      Figure 11. Power Loss


      Being able to measure and compensate for loss in an RF connection is
      important because radios have a receive sensitivity threshold. Sensitivity
      threshold is the point at which a radio can clearly distinguish a signal from
      background noise. The transmitting station must transmit signal that has
      enough amplitude to be recognized by the receiver because receiver’s
      sensitivity is not infinite. Losses between the transmitter and receiver can be
      corrected either by removing the object causing the loss or by increasing the
      transmission power.


c.    Reflection
      Reflection occurs when an electromagnetic wave hits an object with a large
      dimension. Reflections occur on most surfaces and if the surface is smooth, the
      reflected signal will remain unchanged with minimal or reduced absorption and
      scattering of the signal.




D-Link Australia & NZ                                                         Page 13 of 85
Version 1.0
                                                                           Wireless LAN




          Incoming RF                                            Reflected RF




      Figure 12. Reflection


      RF signal reflection can cause serious problem for wireless LAN. The reflection
      of the main signal from many objects in the transmission is called multipath.
      Multipath affects wireless LAN and degrades or even cancels the main signal
      and recreates holes or gaps in the RF coverage area. Examples of multipath
      that cause severe reflection are metal roofs, metal doors, lakes, etc.


d.    Refraction
      Refraction is the bending of a radio wave as it passes through a medium of
      different density. When passing through such medium, some of the radio
      waves will be reflected away and some will be bent through the medium in
      another direction.


                                                                 Reflected RF
            Incoming RF




                                                          Refracted RF

      Figure 13. Refraction


      Refraction can be a serious problem for long distance RF link. As the
      atmosphere changes, the RF waves may change direction and divert the signal
      away from the intended target.


e.    Diffraction



D-Link Australia & NZ                                                     Page 14 of 85
Version 1.0
                                                                          Wireless LAN


      Diffraction occurs when the radio path between the transmitter and receiver is
      obstructed by a rough surface.

                                                              New Wave
                                                              Direction




                 Old Wave                         Building
                 Direction




                    Antenna                                  New Wave
                                                             Direction
                                    Old Wave
                                    Direction

      Figure 14. Diffraction


      Diffraction is the slowing of the wave front at the point where the wave front
      strikes an obstacle, while the rest of the wave front maintains the same speed
      of propagation. It is the effect of waves bending around the obstacle.


f.    Scattering
      Scattering occurs when the medium through which the wave travels consists of
      objects with dimensions that are relatively small compared to the wavelength of
      the signal. Scattered waves are produced by rough surfaces or small objects.
      Scattering can occur when a wave strikes an uneven surface and is reflected in
      many directions simultaneously. There are many small amplitude reflections
      which can destroy the main RF signal. RF signal degradation can cause
      intermittent disruption in communications and signal loss.

                                                               Scattered RF
               Incoming RF




      Figure 15. Scattering



D-Link Australia & NZ                                                     Page 15 of 85
Version 1.0
                                                                            Wireless LAN


      Scattering can also occur as the signal wave travels through particles in the
      medium, such as heavy dust content. Instead of being reflected off an uneven
      surface, the RF waves are individually reflected on very small, tiny particles.


g.    Absorption
      Absorption occurs when the RF signal strikes an object and is absorbed into
      the material of the object. The RF signal does not pass through, reflect off, or
      bend around the object.


                                Incoming RF




                                                         Absorbed RF




      Figure 16. Absorption


Voltage Standing Wave Ratio (VSWR)
VSWR occurs when there is mismatched impedance between devices in an RF
system. Impedance is the resistance to current flow, measured in Ohms. Mismatch
means that an equipment has higher or lower impedance compared to another
equipment connected to it. VSWR is caused by an RF signal reflected at a point of
the impedance mismatch in the signal path. VSWR will result in the loss of forward
energy through a system due to some of the power is being reflected back towards
the transmitter called Return Loss. If the impedances at the ends of a connection are
mismatched, then the antenna will not receive the transmitted power.


For purpose of illustrating VSWR, an example of water flowing through two hoses
may be used. When the two hoses are of the same diameter, water is able to flow
through seamlessly. If the hose connected to a tap is much larger than the second
hose down the other end, there would be a backpressure on the tap and also the
connection between the two hoses.




D-Link Australia & NZ                                                      Page 16 of 85
Version 1.0
                                                                             Wireless LAN




                                     Lower                       Higher
            Backlog                  Impedance                   Impedance
            of water                 Hose                        Hose
            (Return
            Loss)


Figure 17. VSWR – like water through a hose


VSWR is a ratio of impedance mismatched against a perfect impedance match. The
typical VSWR value is 1.5 : 1. The second number is always 1, which represent the
perfect match. The first number will vary. The lower the first number, the better
impedance matching the system is.


Excessive VSWR is a serious problem in an RF circuit. It will result in a decrease in
the amplitude of the transmitted RF signal. When some transmitters are not protected
against power being applied to the transmitter output circuit, the reflected power can
burn out the electronics of the transmitter. When the transmitter circuits is burned out,
VSWR’s effects will appear. The power output levels are unstable and the power
observed is different from the expected power. The methods for changing VSWR in a
circuit include proper use of proper equipment, such as tight connections between
cables and connectors, use of impedance match hardware throughout, and the use
of high-quality equipment with calibration reports.


To prevent negative effects of VSWR, it is important that all cables, connectors and
devices have impedances that match as closely as possible to each other. Most of
today’s wireless LAN devices have impedance of 50 ohms, so 75-ohm cable should
not be used.




Spread Spectrum
Spread spectrum is a communications technique characterized by wide bandwidth
and low peak power. In wireless LAN, modulation techniques used by spread
spectrum communication has many advantages over narrowband communication.
Noise-like spread spectrum signal is hard to detect and hence it is harder to intercept


D-Link Australia & NZ                                                      Page 17 of 85
Version 1.0
                                                                            Wireless LAN


or demodulate without proper tools. In spread spectrum communication, it is lesser
affected by jamming and interference.

       Power                             Narrowband
                                         (High Peak Power)




                                                      Spread Spectrum
                                                      (Low Peak Power)

                                                                      Frequency


Figure 18. Narrowband vs. Spread Spectrum on a frequency domain




Frequency Hopping Spread Spectrum (FHSS)
FHSS is a spread spectrum technique that uses frequency agility to spread data over
more than 83 MHz. Frequency agility refers to the radio’s ability to change
transmission frequency abruptly within the usable RF frequency band. In frequency
hopping wireless LANs, the usable portion of the 2.4 GHz ISM band is 83.5 MHz.


In frequency hopping systems, the carrier changes frequency, or hops according to a
pseudorandom sequence. The sequence is a list of several frequencies that the
carrier will hop at specified time intervals before repeating the pattern. The
transmitter uses this hop sequence to select its transmission frequencies. The carrier
will remain in a certain frequency for a specified time called dwell time. It then uses a
small amount of time, called the hop time, to hop to the next frequency. When the list
of frequencies have completed, the transmitter will repeat the whole sequence again
until the information is received completely.




D-Link Australia & NZ                                                      Page 18 of 85
Version 1.0
                                                                           Wireless LAN



          Transmission
          Frequency
          (GHz)
                                                   Repeating
         2.4835                                    Sequence




         2.4000

                                            Elapsed Time

Figure 19. Single Frequency Hopping System


Frequency hopping is a method of sending data where the transmission and
receiving systems hop along a repeatable pattern of frequencies together. Frequency
hopping systems are resistant to narrow band interference. If a signal interferes with
the frequency hopping signal, only that portion of the spread spectrum signal would
be lost. The rest of the spread spectrum signal would remain intact and the lost data
would be retransmitted.


In reality, an interfering narrow band signal may occupy several megahertz
bandwidth. Since frequency hopping band is over 83 MHz wide, this interfering signal
will cause minimal degradation of the spread spectrum signal.


IEEE 802.11 standard specifies data rates of 1 and 2 Mbps and it must be operating
in the 2.4 GHz ISM band. It allows operation in the range of 2.4000 GHz to 2.4835
GHz.


==Reference Only==
A frequency hopping system will operate using a specified hop pattern called channel.
Some frequency hopping systems allow creating of custom hop patterns and others
allow synchronization between systems to completely eliminate collisions in a co-
located environment.


It is possible to have as many as 79 synchronized, co-located access points.
However, each frequency hopping ratio requires precise synchronization with the
others in order not to interfere with another frequency hopping ratio in the area. To be


D-Link Australia & NZ                                                     Page 19 of 85
Version 1.0
                                                                             Wireless LAN


practical, if synchronized ratios are used, the maximum number of co-located
systems should be 12.


If non-synchronized radios are used in medium-traffic wireless LAN, 26 systems can
co-locate in this wireless LAN. In an environment that has significantly large file
transfer, the practical limit is about 15 co-located systems. More than 15 co-located
frequency hopping systems in such environment will interfere and collisions will begin
to reduce the throughput of the wireless LAN.


The time to transmit on a specific frequency is called dwell time. Once the dwell time
has expired, the system will switch to another frequency and begin to transmit again.


When changing the transmission frequency from one to another, it either switches to
a different circuit tuned to the new frequency or it must change some element of the
current circuit in order to tune to the new frequency. The process of changing to the
new frequency must be completed before transmission can resume. This small
amount of time is called hop time. Hop time is usually not significant compared with
dwell time.


The longer the dwell time, the greater is the throughput. However, there is a limit to
this dwell time. The FCC defines the maximum dwell time of a frequency hopping
spread spectrum system at 400 ms per carrier frequency in any 30 seconds time
frame. Normally, frequency hopping radios are not programmed to operate at a limit.
This is for the operator to have the flexibility for adjustment. By adjusting the dwell
time, the administrator can optimize the frequency hopping spread spectrum network
for the different level of interference. In area where there is little interference, longer
dwell time may be set and hence throughput is higher. In area with interference,
many retransmissions are needed due to corrupted data packets, therefore shorter
dwell time is set.
==End of Reference==


Direct Sequence Spread Spectrum (DSSS)
DSSS is very widely known and the most commonly used spread spectrum type.
This is due to ease of implementation and high data rate. Majority of the wireless
LAN equipment uses DSSS technology. DSSS is a method of sending data which the
transmitting and receiving systems are both on a 22 MHz-wide set of frequencies. A



D-Link Australia & NZ                                                        Page 20 of 85
Version 1.0
                                                                         Wireless LAN


wide channel enables devices to transmit more information at a higher data rate
compared to FHSS systems.


DSSS combines a data signal at the sending station with a higher data rate bit
sequence called chipping code or processing gain. A high processing gain increases
the signal’s resistance to interference. The process of direct sequence begins with a
carrier being modulated with a code sequence. The number of ‘chips’ in the code will
determine how much spreading occurs. The number of chips per bit and the speed of
the code will determine the data rate.


In the 2.4 GHz ISM band, IEEE 802.11 specifies the use of DSSS at a 1 or 2 Mbps.
Under 802.11b standard, the data rates are 5.5 and 11 Mbps. 802.11b is backward
compatible with 802.11 standard.


802.11a standard can operate up to 54 Mbps, uses the 5 GHz UNII (Unlicensed
National Information Infrastructure) bands. It is incompatible with 802.11 and 802.11b.


802.11g is able to operate up to 54 Mbps too. In addition, the advantage of this
standard is that it is backwards compatible with 802.11 and 802.11b.


Unlike frequency hopping systems that use hop sequence to define channels, direct
sequence systems use more conventional definition of channels. Each channel is a
contiguous band of frequencies that is 22 MHz wide. FHSS is using 1MHz carrier
frequencies.

  Power
               Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch
                1  2  3  4  5  6  7  8  9 10 11 12 13




                                                                 Frequency
  2.401                                                                      2.483
  GHz                   3 MHz                 3 MHz                          GHz

Figure 20. DSSS Channel Allocation and Spectral Relationship

D-Link Australia & NZ                                                    Page 21 of 85
Version 1.0
                                                                         Wireless LAN




The above chart is a complete list of channels that are used in most part of Asia and
Europe. 802.11b and 802.11g standard specification specifies 13 channels in this
region. Each frequency is a centre frequency. From this centre, 11 MHz is added and
subtracted to get the useable 22 Mbps wide channel. The channels directly next to
each other overlap significantly.


The table shows the DSSS frequency assignments

 Channel ID      Channel Frequencies (GHz)

 1               2.412

 2               2.417

 3               2.422

 4               2.427

 5               2.432

 6               2.437

 7               2.442

 8               2.447

 9               2.452

 10              2.457

 11              2.467

 12              2.467

 13              2.473


Using DSSS systems with overlapping channels in the same space would cause
interference between systems. DSSS systems with overlapping channels should not
be co-located as it will reduce the wireless LAN performance. This is because the
centre frequencies are 3 MHz apart and each channel is 22 MHz wide, channels
should be co-located only if the channel numbers are at least five channels apart. For
example, channel 1 and 6 do not overlap, channel 6 and 11 do not overlap, etc.
There is a maximum of three co-located direct sequence systems because there are
only 3 non-overlapping channels.



D-Link Australia & NZ                                                    Page 22 of 85
Version 1.0
                                                                          Wireless LAN




        Power
                                    3 MHz


                                        22
                                       MHz



                Ch                     Ch                     Ch
                 1                      6                     11

                                                                   Frequency
     2.401                                                                     2.483
     GHz                                                                       GHz

Figure 21. DSSS Non-Overlapping Channels


Like frequency hopping systems, direct sequence systems are also resistant to
narrow band interference due to their spread spectrum characteristics. A DSSS
signal is more susceptible to narrow band interference than FHSS. This is because
DSSS band (22 MHz wide) is much smaller than FHSS band (79 MHz wide). The
information using DSSS is transmitted along the entire band simultaneously instead
of one frequency at a time. With FHSS, the interference is only influential for a short
time due to the changing of the frequency, which means, only a small portion of the
data may get corruption.




Comparing FHSS and DSSS
Both FHSS and DSSS technologies have their advantages and disadvantages.
There are several factors that need to be considered on which technology is
appropriate.
a.     Narrowband Interference
       FHSS has greater resistance to narrowband interference. DSSS systems may
       be affected by narrowband interference more than FHSS.


b.     Cost
       The cost of implementing a direct sequence system is far lesser than frequency
       hopping system. DSSS equipment is widely available in today’s marketplace.
       This fast adaptation drives the cost to go down.


c.     Co-location


D-Link Australia & NZ                                                     Page 23 of 85
Version 1.0
                                                                         Wireless LAN


      FHSS supports more co-locations compared to DSSS. Since frequency
      hopping systems are always changing frequency and make use of 79 discrete
      channels, it is an advantage of FHSS compared to DSSS with maximum three
      co-located access points.


      However, DSSS has better throughput compared to FHSS. The maximum
      throughput for DSSS is 33 Mbps (3 access points x 11Mbps). For FHSS, the
      maximum throughput is 24 Mbps (12 access points x 2 Mbps).


d.    Equipment Compatibility & Availability
      The Wi-Fi (Wireless Fidelity) Alliance provides testing of 802.11b and 802.11g
      compliant DSSS wireless LAN equipment to ensure interoperability. There are
      no compatibility tests for equipment that uses FHSS. Due to the popularity of
      802.11b and 802.11g products, there is a growing demand for Wi-Fi compliant
      products while the demand for FHSS products is decreasing.


e.    Data Rate and Throughput
      The frequency hopping system is slower than DSSS systems. Most of the
      frequency hopping system’s data rate is only 2 Mbps. Although there are
      systems that operate at more than 3 Mbps, they are not 802.11 compliant.
      DSSS system’s throughput is up to 54 Mbps in 802.11g.


f.    Security
      Frequency hopping systems are less secure than direct sequence systems
      mainly because it is not popular and there is very minimal number of
      manufacturers. These few manufacturers will make use of the standard set of
      hop sequences produced by the standard body so that they can sell their
      products efficiently. Therefore breaking the code of hop sequences is relatively
      simple. Another reason is that the channel number is broadcasted openly with
      each beacon. The MAC address of the transmitted access point can be seen in
      each beacon. For those designs that allow flexible defining custom hopping
      patterns, there is no security because by using spectrum analyzers or standard
      computer, it can track the hopping pattern of a FHSS radio in seconds.




D-Link Australia & NZ                                                    Page 24 of 85
Version 1.0
                                                                              Wireless LAN


Wireless LAN Infrastructure - Super G
Super G from Atheros is a series of intelligent mechanisms that is activated when
additional bandwidth is needed. Its purpose is to increase the actual throughput in a
wireless network. There are four capabilities of Super G that operate independently
to enhance the throughput of wireless LAN in different ways.

                                                                       Throughput
                                                                       (Mbps)
                                                                         60
     Bandwidth      Dynamic Turbo: Dual channels to double rates,
     Benefits       dynamically adjusts for need
                                                                         40
                    Fast Frames & Compression: Packet
                    aggregation & timing modification, standard
                    LempelZiv compression
                    Bursting: More data packets in a given time
                    regardless of the AP type
                                                                         22
                    BaseMode: Standard 802.11 enhanced Tx
                    power and Rx Sensitivity



Figure 22. Super G Technology Suite


a.      Bursting
        Frame bursting is a transmission technique supported by 802.11e draft QoS
        specification. Frame bursting increases throughput of any 802.11a, 802.11b
        and 802.11g link by reducing the overhead associated with the wireless
        transmission. This ability for high data throughput is available for both
        homogeneous and mixed networks.


        Super G has bandwidth enhancements that begin with the frame bursting
        mechanisms. It allows a transmitting device to send multiple frames in a “burst”
        rather than pause after each frame. That means more information is
        transmitted during each transmission opportunity for a given station.


        In a standard transmission, it is separated by the distributed interframe space
        (DIFS). All devices must contend for airtime to transmit their data. After
        transmitting one frame successfully, the devices will contend for the airtime
        again.




D-Link Australia & NZ                                                      Page 25 of 85
Version 1.0
                                                                                               Wireless LAN


      In a burst transmission, the devices only need to contend for the airtime once
      before sending a series of data frames. The overhead of contending for airtime
      between frame transmissions is reduced. There is only short interframe space
      (SIFS) between frames.


                        DIFS &                        DIFS &                         DIFS &
                        Backoff                       Backoff                        Backoff


             Source
                                  Frame 1                       Frame 2
         Without
                                                                                                  Time
         Bursting
                                               ACK                            ACK

               Dest

                                            SIFS                           SIFS


            Source
                        Frame 1                    Frame 2                 Frame 3
          With
                                                                                                  Time
         Bursting
                                      ACK                       ACK                     ACK

              Dest

                                   SIFS     SIFS             SIFS   SIFS             SIFS



      Figure 23. Burst Timing


b.    Fast Frames
      Fast frames enhance data throughput by increasing the number of bits sent per
      data frame via bundling two data frames into a single wireless LAN frame. This
      will eliminate the extra wireless network overhead.


      In a typical network, the maximum frame size for both wired and wireless is
      1500 bytes. Fast frames operate by changing the algorithms that determine the
      structure of the actual data frame and the frame bursting effect. Once fast
      frames are negotiated over a specific wireless link, both the access point and
      wireless client can send wireless frames of 3000 bytes. With the exception of
      Atheros solution, however, most bursting implementations do not provide fast
      frames. This requires an access point that supports fast frames.




D-Link Australia & NZ                                                                          Page 26 of 85
Version 1.0
                                                                                            Wireless LAN



                        DIFS &                       DIFS &                       DIFS &
                        Backoff                      Backoff                      Backoff


            Source
                                  Frame 1                      Frame 2
            Without
             Fast                                                                             Time
            Frames
                                               ACK                          ACK

              Dest

                                            SIFS                         SIFS


                        DIFS &
                        Backoff


            Source
                                    Frame 1 & 2
              With
              Fast                                                                            Time
             Frame
                                                           ACK

              Dest

                                                       SIFS




      Figure 24. Fast Frames


c.    Compression
      A complete hardware data compression engine is embedded in the wireless
      chipsets. The hardware can operate in real-time to enhance throughput for
      many types of network traffic without affecting any of the algorithms used in the
      data transmission or framing technique.


      Super G implements the standards-based Lempel Ziv algorithm that is used in
      popular programs such as PKzip, Winzip, etc. This compression is
      implemented on a per frame basis and affects only data frames. This engine
      compresses before transmission and decompresses after reception. The effect
      of this is the increased of data throughput of the compressed wireless link. It
      also means less airtime is required to transmit each frame.




d.    Dynamic Turbo


D-Link Australia & NZ                                                                       Page 27 of 85
Version 1.0
                                                                           Wireless LAN


      Super G features a multi-channel capability that doubles the effect of all other
      enhancements and increases the range of a data link at any given data rate by
      adapting the way in which radio spectrum is used. This is called Dynamic Turbo.


      Dynamic turbo operates by using the spectrum offered by two radio channels to
      transmit data, just like Ethernet trunking. With multiple radio channels, the data
      rate is doubled. The effective range of a network is increased as the data rate
      of wireless networks decreases the further a station is from access points.


      Dynamic turbo is engaged based on the network traffic ‘demand’ and
      environmental conditions. The access points with dynamic turbo can switch
      dynamically to this mode when an associated wireless client requires the
      greater bandwidth. Third-party wireless clients are not able to perform in
      dynamic turbo mode. The access point will dynamically reconfigure itself for
      multi-channel or single-channel modes, depending on the wireless clients.


The table below shows a summary on the characteristics and benefits for Super G.

    Feature               Characteristics                         Benefits

 Bursting          - More data frames per time        - Overhead is reduced due to
                     period                             increasing in throughput

                   - Based on standards               - Subset of 802.11e

                   - Relevant to STA                  - Can be applied to any access
                                                        point

 Fast Frames       - Utilizes frame aggregation -         Transmit more data per
                     and timing modifications             frame and hence increase
                                                          the throughput

 Compression       - Real-time hardware        data -     Compressed       data      can
                     compression                          increase data

                   - Lempel Ziv standard              -   No impact on host processor

 Dynamic           - Similar    to  trunking    in -      Multiple   channels    will
 Turbo               Ethernet, uses dual channels         maximize the bandwidth
                     to “double” the transmission
                     rates                         -      Aware of the environment

                   - Adjust the bandwidth after
                     analyzing the environment




D-Link Australia & NZ                                                     Page 28 of 85
Version 1.0
                                                                            Wireless LAN


Wireless Distribution System (WDS)
In IEEE 802.11, a distribution system is a system that interconnects Basic Service
Sets (BSS). A BSS is a cell in which an access point can cover. A distribution system
connects these cells together to build a bigger area network that allows mobile users
to roam and stay connected to the network resources using the wireless equipment.


In a typical Wired Distribution System, the access points in the same network are
connected together using cable.




                                                            Access
                                                            Point
                           BSS
                        Channel 6
                                        Wired
                                     Distribution
                                       System
                                                        BSS
                                    Access
                                    Point            Channel 11




Figure 25. Typical Wired Distribution System


However if no cable is used in the distributed system, the connection between the
access points can be established using wireless modules. This single wireless
module in the access point can perform multiple roles at the same time. It can
connect wireless clients to the infrastructure and it can maintain up to six different
wireless connections to other access points. The access point will denote “port 1” as
its Ethernet port and “port 2 to 7” for all the six different wireless connections. It is
necessary for the operational frequency channel to be the same for the cell that is
controlled by the access point and for the wireless links to other access points.




D-Link Australia & NZ                                                      Page 29 of 85
Version 1.0
                                                                           Wireless LAN




                                                                  Access
                                                                  Point
                           BSS
                        Channel 1
                                         Wireless
                                       Distribution
                                         System
                                                            BSS
                                     Access
                                     Point                Channel 1




Figure 26. Wireless Distribution System


There is a wireless module in the access point and it contains Media Access Control
(MAC) address. The wireless client with the wireless module also has MAC address.
In a WDS link, four MAC addresses are involved. That includes sender and
destination computers, sender and destination access points. All these MAC
addresses are included in the 802.11 frame.


Upon receiving the 802.11 frame, the wireless module in the access point will convert
it to an 802.3 Ethernet frame. The 802.3 Ethernet frame consists of both the source
and destination computer’s MAC addresses. It will also pass the frame to the bridge
address table. This bridge address table consists of all the wired and wireless
computers connecting directly or indirect to it. The wired computers will be listed as
“port 1”. As for wireless computers, it will be listed as one of the six wireless LANs
that is associating to the access point as port 2 to 7.




D-Link Australia & NZ                                                      Page 30 of 85
Version 1.0
                                                                                    Wireless LAN




      Computer A                                                                Computer B
                             Access Point A           Access Point B


           Send frame with MAC
           addr of Computer B
           Acknowledge receive
           the traffic

                        The forwarding table
                        indicates that the
                        frame is to pass to its
                        own PC card
                        PC card look up its
                        own forwarding table
                                     Forward the frame to
                                     the PC card of AP B
                                     Acknowledge receive
                                     the traffic

                                                  The forwarding table
                                                  indicates that the
                                                  frame is to pass to
                                                  Computer B
                                                              Forward the frame to
                                                              the computer

                                                              Acknowledge
                                                              receive the traffic


Figure 27. Steps in Traffic Flow in WDS


Roaming between cells that are interconnected by a WDS link works exactly the
same as the cells that are interconnected via Ethernet. The bridge learning table will
be upgraded when there is a relocation of a wireless client from one cell to another.
The Inter Access Point Protocol (IAPP) handles the hand-over request messages.


Due to the flexibility of WDS, there are a few configurations that can be implemented.
a.    Star Configuration
      Star configuration can cover a more rectangular or square area. The central
      access point is connected to the wired infrastructure network.



D-Link Australia & NZ                                                               Page 31 of 85
Version 1.0
                                                                       Wireless LAN




      Figure 28. Star Configuration using WDS


b.    Chain Configuration
      A chain configuration allows coverage for a longer shape. The first access is
      connected to the wired infrastructure network.




D-Link Australia & NZ                                                 Page 32 of 85
Version 1.0
                                                                       Wireless LAN




      Figure 29. Chain Configuration in WDS


c.    Ring Configuration
      When the end points of a chain are connected to each other, a loop is created.
      It is advisable to avoid ring configuration because it will lead to bad
      performance, broadcast and multicast storms. If the access points support
      spanning tree protocols, the ring will allow redundancy in case one of the
      access points fails.




D-Link Australia & NZ                                                  Page 33 of 85
Version 1.0
                                                                                   Wireless LAN




                                       STP
                                      Enabled




      Figure 30. Ring Configuration in WDS


Wireless Distribution System (WDS) offers a great flexibility at low cost and it can be
applied in many situations. However, there are a few considerations to make before
deciding to use WDS.
a.    Advantages of WDS
      Without additional cost, the existing access point with WDS function can have a
      WDS link by reconfiguring the device. There is no need to pay for an additional
      wireless module.


      It is more flexible when adding a wireless point compared to a wired Ethernet
      point. WDS is able to create a roaming network without the hazard of installing
      physical cables. It is excellent for areas where cables are not accessible.


b.    Disadvantages of WDS
      It is not possible to use encryption with dynamic assigned rotating keys on a
      WDS link*. Only fixed assigned Wired Equivalent Privacy (WEP) keys can be
      used. If the user wants to secure their network by using 802.1x, it will not be
      able to use WDS. *except certain specific models like DWL-3200AP   which supports WPA/WPS2-

      PSK


D-Link Australia & NZ                                                             Page 34 of 85
Version 1.0
                                                                             Wireless LAN




Basic Principles of Antennas
We will discuss on the basic antenna principles that directly relates to the use of
wireless LAN. It is important for network administrator to understand the antenna
design very well in order to administrate the network.


The following are two important key points to understand about antennas.
      Transmitting antennas convert electrical energy into RF waves. Receiving
      antennas converts RF waves into electrical energy.
      Physical dimensions of an antenna, such as its length, are directly related to
      the frequency at which the antenna can propagate waves or receive
      propagated waves.




Type of Antennas
An RF antenna is a device used to convert high frequency signals on a transmission
line (for example, cable) into propagated waves in the air. The electrical fields
emitted from antennas are called beams or lobes.


There are three generic categories of RF antennas. Each category has different RF
characteristics and appropriate usages.
a.    Omni-directional (Dipole) Antennas
      Omni-directional antenna is the most common wireless LAN antenna. It is
      simple to design and it becomes the standard antenna for most access points.
      It radiates its energy equally in all directions around its axis. It concentrates its
      energy into a cone, known as “beam”.




                  Dipole Doughnut                        Dipole Side view



      Figure 31. Omni-Directional (Dipole) Antenna


D-Link Australia & NZ                                                        Page 35 of 85
Version 1.0
                                                                         Wireless LAN




      The radiated energy for a dipole is concentrated into a region that looks like a
      doughnut. In the above figure, the dipole is positioned vertically through the
      “hole” of the doughnut. The signal from the omni-directional antenna radiates in
      a 360-degree horizontal beam equally. In the side view, the dipole antenna will
      form a “figure 8”.




                           Side View
                                                           Top View


      Figure 32. Coverage Area of an Omni-Directional Antenna


      High gain omni-directional antennas offer more horizontal area but the vertical
      coverage area is reduced.




                            Side View


                                                                 Top View


      Figure 33. Coverage Area of a High-Gain Omni-Directional Antenna


      Omni-directional antennas are used when coverage in all directions around the
      horizontal axis of the antenna is required. They are most effective when large
      coverage of areas are needed around a central point.


b.    Semi-directional Antennas



D-Link Australia & NZ                                                    Page 36 of 85
Version 1.0
                                                                            Wireless LAN


      Semi-directional antennas come in many different styles and shapes. The
      frequently used types are Patch, Panel and Yagi antennas. They have different
      coverage characteristics. These antennas direct the energy from the transmitter
      significantly more in one particular direction. They often radiate in a
      hemispherical or cylindrical coverage pattern.




             Directional Patch Antenna
                                                 Directional Yagi Antenna


      Figure 34. Coverage Area of a Semi-Directional Antenna


      Semi-directional antennas are suitable for short and medium range bridging. It
      is also ideal for using indoors where the antennas are mounted at one side of
      the whole coverage area.


c.    Highly-directional Antennas
      Highly-directional antennas emit the most narrow signal beam of any antenna
      type and have the greatest gain of the three groups of antennas. They are
      typically concave, disk-shaped devices. These antennas are ideal for long
      distance, point-to-point wireless links.




      Figure 35. Radiation Pattern of a Highly-directional Antenna


The RF antenna concepts and fundamentals are needed to allow the administrator to
understand how wireless LAN equipment functions over the wireless medium. A solid
understanding of the basic antenna functionality is important. It includes how to
position the antennas, how much power they radiate, the distance the radiated power
is likely to travel and how much the power can be picked up by the receiver, etc.


a.    Line of Sight (LOS)

D-Link Australia & NZ                                                       Page 37 of 85
Version 1.0
                                                                            Wireless LAN


      LOS is defined as a straight line from the object in sight (the transmitter) to the
      observer’s eye (the receiver). However, LOS is not exactly a straight line
      because light waves can change direction due to refraction, diffraction and
      reflection. RF works in a similar way as visible light within wireless LAN
      frequencies, but RF LOS can also be affected by blockage of the Fresnel Zone.

                                        Line of Sight




      Figure 36. Line of Sight


b.    Fresnel Zone (pronounced as “fra-NEL”)
      The Fresnel Zone occupies a series of concentric ellipsoid-shaped area around
      the LOS path. The Fresnel Zone is important to the integrity of the RF link
      because it defines an area around the LOS that can introduce RF signal
      interference if blocked. Examples of objects in the Fresnel Zone are trees,
      hilltops and buildings that can diffract or reflect the main signal away from the
      receiver, changing the RF LOS. These objects can absorb or scatter the main
      RF signal, causing degradation or complete signal loss.



                                        Fresnel Zone




      Figure 37. Fresnel Zone


      The radius of the Fresnel Zone at its widest point can be calculated by the
      following formula.

D-Link Australia & NZ                                                      Page 38 of 85
Version 1.0
                                                                             Wireless LAN




                                                     d
                            r = 43.3 x
                                                     4f
      where d is the link distance in miles, f is the frequency in GHz, and the answer r,
      is in feet. For example, suppose there is a 2.4 GHz link 5 miles in length. The
      resulting Fresnel Zone would have a radius of 31.25 feet.


      Some blockage of the Fresnel Zone can occur without significant link disruption.
      To be conservative, it is suggested to allow no more than 20% blockage of the
      Fresnel Zone.


      Fresnel Zone cannot be defined clearly in indoor installations unless the signal
      is partially or fully blocked. In most indoor installations, RF signals pass through,
      reflect off and refract around the walls, furniture and other obstructions. Most of
      the wireless users are mobile and the Fresnel Zone is constantly changing.


c.    Polarization
      The radio wave is made of two fields, one electric and one magnetic. These
      two fields are on planes that are perpendicular to each other.

                                                               H-Plane


                                                                         E-Plane


                Antenna




      Figure 38. E-planes and H-planes


      The sum of the two fields is called the electro-magnetic field. “Oscillation” is the
      process where energy is transferred back and forth from one field to the other.
      The plane that is parallel with the antenna is called the “E-plane” and the plane
      that is perpendicular to the antenna is called the “H-plane”. The wave



D-Link Australia & NZ                                                        Page 39 of 85
Version 1.0
                                                                              Wireless LAN


      polarization is the position and direction of the electric field with reference to the
      Earth’s surface.


      Polarization is the physical orientation of the antenna in a horizontal or vertical
      position. The electric field that is parallel to the ground is called horizontal
      polarization while the electric field that is perpendicular to the ground is called
      the vertical polarization.


      Vertical polarization is typically used in wireless LANs and it is perpendicular to
      the Earth’s plane. In most access points, the antennas are sticking up vertically.
      Antennas that are not polarized in the same way are not able to communicate
      with each other effectively.


d.    Antennas Gain
      An antenna is a passive device without any amplifiers and filters associating
      with it. The antenna does not condition, amplify or manipulate the signal.
      Antenna amplification is the result of focusing the RF radiation into a tighter
      beam. The focusing of the radiation is measured by ways of horizontal and
      vertical beamwidth in degrees.


      Omni-Directional antenna has a 360-degree horizontal beamwidth. By limiting
      the 360-degree beamwidth into a more focused beam, say 30 degrees, at the
      same power, the RF waves will be radiated further.




D-Link Australia & NZ                                                         Page 40 of 85
Version 1.0
                                                                           Wireless LAN




                               a

                                                        a

                                                      Vertical Beamwidth
                Horizontal Beamwidth

                                   Omni-Directional Antenna




                                                      >a


                                       Yagi Antenna

      Figure 39. Different distance using different antenna with same power


e.    Beamwidth
      Focusing antenna beams can increase the antenna’s gain. The antenna’s
      beamwidth means the ‘width’ of the RF signal beam that the antenna transmits.




                                                           Horizontal
                                                           Beamwidth




                                                      Vertical
                                                      Beamwidth


      Figure 40. Beamwidth of an Antenna


      The two vectors for an antenna are vertical and horizontal beamwidth. The
      vertical beamwidth is perpendicular to the Earth’s surface and the horizontal

D-Link Australia & NZ                                                      Page 41 of 85
Version 1.0
                                                                          Wireless LAN


      beamwidth is parallel to the Earth’s surface. Both vectors are measured in
      degrees. Each type of antenna has different beamwidth specifications. The
      chart below shows a general guideline for the beamwidth.



       Antenna Type            Horizontal Beamwidth          Vertical Beamwidth

       Omni-Directional        360                           7 to 80

       Patch / Panel           30 to 180                     6 to 90

       Yagi                    30 to 78                      14 to 64

       Parabolic Dish          4 to 25                       4 to 21


      Selecting an antenna with the appropriate wide or narrow beamwidths is
      essential in having the desired RF coverage pattern.


f.    Free Space Path Loss
      Free space path loss refers to the loss incurred by an RF signal due to “signal
      dispersion” which is a natural broadening of the wave front. The wider the wave
      front, the less power can be induced into the receiving antenna. As the
      transmitted signal propagates, its power level decreases at a rate inversely
      proportional to the distance traveled and proportional to the wavelength of the
      signal. This power level is an important factor when considering link viability.
      Path loss represents the single greatest source of loss in the wireless system.


g.    Intentional Radiator
      An intentional radiator is an RF device that is specifically designed to generate
      and radiate RF signals. This intentional radiator will include the RF device and
      all cabling and connectors up to, but not including, the antenna.




D-Link Australia & NZ                                                     Page 42 of 85
Version 1.0
                                                                           Wireless LAN




                            Power Output of
                            the Intentional
                            Radiator                   Antenna


                                                       Connector


                        Connector
                                                                    Components
                                                                    included in the
              RF device                               Connector     Intentional
                                                                    Radiator


      Figure 41. Intentional Radiator


      Any reference to the “power output of the Intentional Radiator” is referred to the
      power output at the end of the last cable or connector before the antenna.


h.    Equivalent Isotropically Radiated Power (EIRP)
      EIRP is the power actually radiated by the antenna element. This is very
      important because it is used in calculating whether or not a wireless link is
      viable. EIRP will take into account the gain of antenna.


                        EIRP
                        (Output Power)
                                                       Antenna

                                                       Connector


                        Connector
                                                                    Components
                                                                    included in the
              RF device                               Connector     Intentional
                                                                    Radiator



      Figure 42. EIRP


      For example, use an 18 dBi antenna with a 50 cm cable.


      The calculated EIRP will be as follows:



              EIRP = Antenna Gain – Loss via cable
D-Link Australia & NZ                                                      Page 43 of 85
Version 1.0
                                                                         Wireless LAN


      EIRP = 18 dBi – 0.415 dB = 17.585 dB
      (1 meter of cable loss is 0.83 dB)




Antenna Installation
It is very important to install the antennas properly in Wireless LAN. An improper
installation can lead to damage or destruction of equipment and also personal injury.
a.    Placement
      For omni-directional antennas that are attached to the access points, they are
      placed in the middle of the desired coverage area. Placing the antenna as high
      as possible will increase the coverage area too, especially for high-gain omni-
      directional antennas. Outdoor antennas should be mounted above all possible
      obstructions such as tree and building such that there are no objects on the
      Fresnel Zone.


b.    Mounting
      The antenna must be mounted. There is no definite answer of where to mount
      the antenna. A site survey is needed to determine this. Each antenna comes
      with mounting instructions that show how to install and secure the antenna from
      the manufacturers. Also, the antenna will be packaged with its own mounting kit.


      There are some issues that need to be taken into consideration while mounting
      antennas. In certain scenarios, the brackets that are packaged together with
      the antenna may not be suitable. Modifying the brackets or custom making
      another set may be necessary.


      The mounting must be solid and secure and not just hang on by its cable. The
      cable may break and swaying of the cable can produce a moving cell.


      Antennas can be unsightly and is therefore normally hidden. Some
      manufacturers produce ceiling-mount panel antennas. When these aesthetics
      are important, patch or panel antennas are used instead of omni-directional
      antennas.


c.    Appropriate Use




D-Link Australia & NZ                                                    Page 44 of 85
Version 1.0
                                                                         Wireless LAN


      The indoor antennas are used inside the buildings while outdoor antennas are
      used outside the buildings. In scenarios where there is a significantly large
      indoor area, the outdoor antenna may be used.


      Outdoor antennas are mostly sealed to prevent water from entering the
      antenna internal hardware component. They are made of plastics so that they
      can withstand extreme heat and cold. Indoor antennas are not made for
      outdoor usage without the above elements.


d.    Orientation
      Antenna orientation determines polarization which has a significant impact on
      the signal reception. The antenna of the access points and the wireless clients
      should have the same orientation for maximum reception. The throughput of
      the link will be reduced drastically if each end of the link does not have the
      same antenna orientation.


e.    Alignment
      Antenna alignment can be critical in certain scenarios and may not in others.
      Antennas with very wide horizontal and vertical beamwidth allow the
      administrator to easily aim two antennas in a building-to-building bridging
      environment in each other’s general direction and will get an almost perfect
      reception.


      Alignment is more important when implementing long-distance bridging links
      using highly-directional antennas. Most of the wireless bridges come with
      alignment software to help the administrator in optimizing antenna alignment for
      the best reception, which will reduce lost packets and high retry counts while
      maximizing signal strength.


      When using access points with omni-directional or semi-directional antennas,
      proper alignment is needed to cover the appropriate area where the wireless
      clients are located.


f.    Safety
      RF antennas can be dangerous to implement and operate. Always read the
      instruction manual provided by the manufacturers carefully. Following the
      provided instructions will prevent damage to the antenna and personal injury.

D-Link Australia & NZ                                                    Page 45 of 85
Version 1.0
                                                                         Wireless LAN




      Never touch a high-gain antenna to any part of the body or point it towards the
      body while it is transmitting. The transmission power is equivalent to putting
      your body in a microwave oven.


      It is advisable to engage a professional installer to install the wireless LAN.
      These installers are trained with proper climbing safety and will be able to
      provide a better installation and secure the antenna if it is to be mounted on a
      pole, tower or other type of elevated constructions.


      Always keep antennas away from metal obstructions, such as heating and air-
      conditioning ducts, major cabling, etc. These metal obstructions create a
      significant amount of multipath. They can also reflect a large portion of the RF
      signal and this reflected signal can be dangerous to bystanders.


      Antenna towers should keep a safe distance from the overhead power lines.
      The recommended safe distance is twice the antenna height. The antennas
      should not be placed near the power source because an electrical shock
      between the power source and the wireless LAN can be dangerous to people
      working on the wireless LAN and is likely to destroy the wireless LAN
      equipment.




Power-over-Ethernet (PoE)
Power-over-Ethernet (PoE) is a method of delivering DC voltage to an access point
or wireless bridge over the Cat 5 Ethernet cable for the purpose of powering the unit.
PoE is used when the AC power supply is not available at the location where the
wireless LAN infrastructure devices are installed. The Ethernet cable is used to carry
both the power and the data to the units.




D-Link Australia & NZ                                                    Page 46 of 85
Version 1.0
                                                                           Wireless LAN



                Passive Hub or
                Switch



                         Unpowered
                                                                   Access Point
                         Ethernet


                 PoE                                    PoE
                Device                                 Device
                                     Powered
                                     Ethernet
                                                              Unpowered
                                                               Ethernet


Figure 43. PoE Installation Example


Although configuration and management is generally not necessary for PoE device,
there are a few points that need to be taken notice before implementing PoE. There
is a PoE industry standard, 802.3af. However, there are still some minor
manufacturers who do not use this standard. This means that PoE devices from
different manufacturers may not work together. The output voltage required to power
a wireless LAN device may differ from manufacturer to manufacturer if it is not
802.3af compliant.


There are the two common PoE options.
a.    Single-port DC Voltage Injectors
      A pair of single-port DC voltage injectors is needed to connect a set of wireless
      LAN infrastructure device, such as an access point. One of the devices is
      connected to the passive switch and the power socket. The output UTP cable
      contains the powered Ethernet. The other device is connected to the power and
      LAN input of the access point.


b.    Active Ethernet Switches
      For an enterprise installation of access points, an active Ethernet switch is used.
      These devices incorporate DC voltage injection into the Ethernet switch itself
      and allow a large number of PoE devices without any additional hardware in
      the network.




D-Link Australia & NZ                                                      Page 47 of 85
Version 1.0
                                                                               Wireless LAN


Troubleshooting Wireless LAN
In the implementation of wireless LAN, dealing with the behavior of RF signals is the
main challenge. In order to implement a successful wireless LAN, several obstacles
need to be taken into consideration on how to troubleshoot them.
a.    Multipath
      The RF signal grows wider as it is transmitted farther. When the RF signal
      encounters objects in its path, it will reflect, diffract, or interfere with the actual
      signal. Some of the reflected waves will also head towards the receiver. This
      behavior is called multipath. Multipath is defined as the composition of the
      primary signal plus the duplicate or echoed wave cased by the reflections of the
      waves off objects between the transmitter and the receiver. The delay between
      the time that the primary signal arrives and the time that the last reflected signal
      arrives is known as delay spread.


             Transmitter                                                      Receiver




                                                River




      Figure 44. Multipath


      Multipath can cause several conditions which can affect the transmission of the
      RF signal differently.
      -   Decreased Signal Amplitude
          When the RF wave arrives at the receiver, many reflected waves may arrive
          at the same time from different directions. The amplitude for these waves
          are combined and added to the main RF wave. Reflected wave, if out-of-
          phase with the main wave, can decrease signal amplitude at the receiver.
          This occurrence is called downfade.




D-Link Australia & NZ                                                          Page 48 of 85
Version 1.0
                                                                             Wireless LAN



                                       Amplitude decrease
               Antenna                                                       Antenna

                                               Out-of-phase reflected
                                               signal is added to the
                                               main signal




                                          Reflected Surface


          Figure 45. Downfade


      -   Corruption
          Corrupted signals due to multipath can occur as a result of the same
          phenomena that cause decreased amplitude, but it is more serious. When
          the reflected waves arrive at the receiver out-of-phase with the main wave,
          they will cause the wave to reduce in amplitude greatly. Although there is a
          reduction in amplitude, the receiver is sensitive enough to detect most if the
          information carried on the wave, but not all of them.


          The signal to noise ratio (SNR) is generally very low. The receiver is unable
          to decipher between the signal and noise. Therefore, data is only part of the
          transmitted data. This corruption of data will require the transmitter to resend
          the data. It will lead to increasing of overhead and decreasing in throughput
          in the wireless LAN.




D-Link Australia & NZ                                                       Page 49 of 85
Version 1.0
                                                                        Wireless LAN



                                       Reflective Surface



                                         Received signal is
                                         corrupted by
                                         reflected signals
                 Antenna                                            Antenna




                                       Reflective Surface


          Figure 46. RF Signal Corruption


      -   Nulling
          Nulling occurs when one or more reflected waves arrive at the receiver out-
          of-phase with the main wave. This will lead to the main wave’s amplitude
          being cancelled or ‘null’ the entire set of RF waves.


          When nulling occurs, retransmission of the data will not solve the problem.
          The transmitter, receiver, or reflective objects must be moved to
          compensate for the nulling effects on the RF wave.


                                       Reflective Surface


                                      Reflected signals added
                                      to main signal cancel all
                                      signal amplitude and
                 Antenna              result in no signal           Antenna




                                       Reflective Surface


D-Link Australia & NZ                                                   Page 50 of 85
Version 1.0
                                                                             Wireless LAN


          Figure 47. RF Signal Nulling


      -   Increased Signal Amplitude
          Multipath conditions can also cause a signal’s amplitude to be increased.
          Upfade is the term used to describe when multipath causes an RF signal to
          gain strength. The reflected signals will arrive at the receiver in-phase with
          the main signal. All these waves are added to the main signal.


          However, the received signal will not be stronger than the transmitted signal
          due to free path loss. Path loss is the effect of ‘s’ signal losing amplitude as
          the signal travels through an open space. The total signal that reaches the
          receiver will be stronger than the signal without the presence of multipath.

                                    Amplitude decrease
                                    due to Path Loss
               Antenna                                                Antenna

                                             In-phase reflected
                                             signal is added to
                                             the main signal




                                          Reflective Surface


          Figure 48. Upfade


      We cannot see an in-phase or out-of-phase RF wave. We can only look for the
      effects of the multipath to detect its occurrence. A common method of finding
      multipath is to look for the RF coverage holes in a site survey. These holes are
      created due to lack of coverage and also the multipath reflections that cancel
      the main signal. The administrator needs to understand the sources of the
      multipath to eliminate its effects.


      Multipath is caused by reflected RF waves. Obstacles that can easily reflect RF
      waves include metal and water. They should be removed from or avoided in the
      signal path. It also includes the moving of transmitting and receiving antennas.
      Users may roam into an area with high multipath, without knowing why their RF
      signal is upgraded significantly.


D-Link Australia & NZ                                                       Page 51 of 85
Version 1.0
                                                                            Wireless LAN




      Antenna diversity was devised to compensate for multipath. It uses multiple
      antennas, inputs and receivers in order to compensate for the conditions that
      cause multipath. One of the types of receiving diversity is called antenna
      switching diversity. Transmission diversity is commonly used by most wireless
      LAN manufacturer.



       Antenna Diversity        - Multiple antennas on single input
       (not active)
                                - Rarely used

       Switching Diversity      - Multiple antennas on multiple receivers

                                - Switches receivers based on signal strength

       Antenna Switching        - Used by most WLAN manufacturers
       Diversity (active)
                                - Multiple antennas on multiple inputs – single receiver

                                - Signal is received through only one antenna at a time

       Phase Diversity          - Patented proprietary technology

                                - Adjust phase of antenna to the phase of the signal in
                                  order to maintain signal quality

       Transmission             - Used by most WLAN manufacturers
       Diversity
                                - Transmits out of the antenna last used for reception

                                - Can alternate antennas for transmission retries

                                - A unit can either transmit or receive, but not both
                                  simultaneously


      Most of the access points in today’s wireless LAN are built with dual antennas
      so as to compensate for the degrading effects of multipath on signal quality and
      throughput.


b.    Hidden Node
      Collision is always a problem for computer networks. Collisions occur when two
      or   more     nodes    sharing   the   communication    medium     transmit   data
      simultaneously. The two signals will corrupt each other and result in unreadable
      packets. CSMA/CD is used with Ethernet to check the channel before
      transmitting data. It involves checking of the voltage on the wire before


D-Link Australia & NZ                                                       Page 52 of 85
Version 1.0
                                                                               Wireless LAN


      transmitting. However, the process is more difficult for wireless system because
      collisions are undetectable.


      Hidden node is a situation encountered with wireless LANs in which at least
      one node is unable to detect one or more of the other nodes connected to the
      wireless LAN. A node can see the access point but cannot see that there are
      other clients connecting to the same access point. It can be due to obstacles or
      long distance between nodes. This will cause a problem in the sharing of the
      medium, and hence causing collision between node transmissions. It will
      significantly result in degrading the throughput in wireless LAN.


                                                   Access Point




                  Wireless Client A                        Wireless Client B


                                            Wall


      Figure 49. Hidden Node


      The symptom of degradation in the throughput of a wireless LAN is called a
      hidden node. The administrator will normally discover there is a hidden node
      when there is complain on a sudden sluggishness in the network.


      Due to the mobility of wireless LAN, hidden nodes may appear anytime. It is
      therefore necessary to locate the hidden nodes. This process usually includes
      a manual search for the nodes using trial and error. Once the nodes are
      located, there are a few remedies and workarounds for the problem.


      -   Use RTS/CTS
          The RTS/CTS (request-to-send/clear-to-send) protocol may not be a
          solution to the hidden node problem. However, it is a method to reduce the



D-Link Australia & NZ                                                      Page 53 of 85
Version 1.0
                                                                            Wireless LAN


          negative impact of hidden nodes on the network. Hidden node cause
          excessive collisions and has an impact on the network throughput.


          The RTS/CTS involves sending a small RTS packet to the intended recipient
          and prompt it to send back a CTS packet to clear the medium for data
          transmission before sending the data. This process will inform all the nearby
          wireless stations that there is data about to be sent, hence inform them to
          delay transmissions. This will avoid collisions. Both the RTS and CTS
          packets contain the length of the intended data for transmission so that the
          other hearing stations will know how long is the transmission and when they
          can start to transmit again.


      -   Increase Power to the Nodes
          Increasing the power of the nodes can solve the hidden node problem by
          allowing the size of the cell around each node to increase. This will enable
          the normal nodes to detect and hear any hidden nodes. When the normal
          nodes hear the hidden nodes, the hidden nodes are no longer hidden nodes.


      -   Remove Obstacles
          If increasing power on the nodes does not solve the problem, this means
          that the hidden node is located in a cement or steel wall that prevents
          communication with other nodes. Removing the obstacle will allow
          unblocked communication.


      -   Move the Node
          If the obstacles cannot be removed, the administrator might need to force
          the user to move to another area. Alternatively, the administrator can add an
          access point in the hidden area for the proper coverage.


c.    Near / Far
      The near/far problem in wireless LAN implementation may occur if the wireless
      client is in the following scenarios:
      - very near to the access point / having very high power settings
      - very far away from the access point / using much less transmitting power
      This will lead to the wireless clients that are far away from the access point fail
      to be heard due to the “louder” signal by the closer and high-powered clients.



D-Link Australia & NZ                                                      Page 54 of 85
Version 1.0
                                                                             Wireless LAN




                        Access Point

                                                   Unheard Signal




                                       100 mW                                  5 mW

                        3 meters
                                            90 meters




      Figure 50. Near / Far


      In wireless LAN, the node that is within the normal range of the access point is
      being drown out and it fails to hear the signals of the other further clients. The
      administrator has to be aware of this near/far problem in the site survey.


      Troubleshooting the near/far problem can be done by taking a good look at the
      layout with information on the locations of stations. The administrator can also
      use a wireless protocol analyzer to pick up the transmissions from all stations.
      If the node is not heard nor has faint signal, this node is too far.


      The near/far problem can be resolved by increasing the power of the far-end
      nodes, decreasing the power of the near-end nodes, or moving the far-end
      nodes closer to the access points.


d.    System Throughput
      The throughput of wireless LAN is based on many factors. This includes the
      amount and type of interference that may affect the amount of data that can be
      transmitted successfully. Additional security solutions that involve encrypting
      and decrypting of data, such as WEP, can cause a decrease in throughput.
      Using VPN tunnels can add overheads to the wireless LAN system as well.


      When the distance between the transmitter and receiver is far apart, it will
      cause the throughput to decrease due to increase in the number of errors.


D-Link Australia & NZ                                                        Page 55 of 85
Version 1.0
                                                                            Wireless LAN


      Retransmission is needed. In today’s popular wireless LAN, spread spectrum
      systems can make discrete jumps to specified data rate, such as 11, 5.5, 2 and
      1 Mbps. For example, if device cannot maintain at 11 Mbps, it will drop to 5.5
      Mbps. The throughput is about 50% of the data rate on wireless LAN system.
      Changing the data rate will result in a great impact on the throughput.


      Hardware limitation can also dictate the data rate. When an IEEE 802.11b
      device communicates with the IEEE 802.11g device, the maximum data rate is
      only 11 Mbps, even though 802.11g can communicate up to 54Mbps. The
      actual throughput in this case is less than 50%. Another hardware limitation is
      the CPU power of the access point. If the slow CPU cannot handle full 54 Mbps
      and 128-bit WEP, it will affect the throughput.


      Frequency Hopping Spread Spectrum (FHSS) and Direct Sequence Spread
      Spectrum (DSSS) make a difference in throughput too. FHSS is typically
      transmitted as 800 kbps or 1.6 Mbps while DSSS can support up to 11 or 54
      Mbps.


      Other factors that limit the throughput of wireless LAN include proprietary data
      link layer protocols, fragmentation, and packet size. Larger packets can have
      greater throughput because the ratio of data to overhead is better.


      RTS/CTS is used in some of the wireless LAN implementations and it will
      create significant overhead due to the amount of handshaking that is taking
      place during the transfer.


      The number of users attempting to access a medium simultaneously will have
      an impact. The increase of simultaneous users will decrease the throughput of
      each wireless client received from the access point.


      Co-location is a common wireless LAN implementation technique that is used
      to provide more bandwidth and throughput to the wireless users in the given
      area. In wireless LAN, it allows three non-overlapping RF channels (1, 6 and
      11). These three channels can be used to co-locate multiple access points
      within the same physical area using 802.11b/g equipment.




D-Link Australia & NZ                                                       Page 56 of 85
Version 1.0
                                                                               Wireless LAN




                                    Channel 6




                        Channel 1
                                                    Channel 11




      Figure 51. Co-location Throughput


      In reality, however, there is still a small amount of overlap for channel 1 and 6
      (or channel 6 and 11). This overlap is due to the transmission of two access
      points at the same high output power which are located relatively very close to
      each other. Instead of the normal half-duplex throughput from each access
      point, a detrimental effect is seen on all three of them. The throughput can be
      decreased on all three access points.

            Signal                  Channel               Channel
            Level                   Overlap               Overlap




                        Channel 1             Channel 6         Channel 11

                                                                             Frequency

      Figure 52. DSSS Channel Overlap


      Although the above shows overlapping channels using channel 1, 6, and 11,
      and there should not have been three co-located access points, you may still
      use these channels. When you experience degraded throughput, you may keep
      this in mind and change the channels accordingly. You may use two co-located
      access points instead of three.




D-Link Australia & NZ                                                          Page 57 of 85
Version 1.0
                                                                            Wireless LAN




                                                Remove this access point allowing
           Signal                               more channel separation for greater
           Level                                throughput




                        Channel 1        Channel 6         Channel 11

                                                                        Frequency

      Figure 53. Using Two Access Points instead of Three


      Another alternative is to use 802.11a compliant equipment which operates in
      the 5 GHz UNII bands. This 5 GHz UNII bands is wider compared to 2.4 GHz
      ISM band. Having a mixture of 802.11b/g and 802.11a equipment co-located in
      the same space will not have any interference between systems. There are up
      to two (or three) 802.11b/g systems and eight 802.11a systems in the same
      physical space. However, 802.11a equipment has lesser availability and they
      are more expensive compared to the popular 802.11b/g devices.


e.    Type of Interference
      RF technology has many unpredictable behaviors and therefore it is necessary
      to consider many kinds of RF interference when implementing and managing a
      wireless LAN.
      -   Narrowband
          Narrowband RF is the opposite of spread spectrum technology. Narrowband
          signals can interrupt the RF signals from a spread spectrum device such as
          access points. Narrowband signal primarily disrupt the RF signals in channel
          3, as such, if channel 11 is used, there may not such interference
          experienced. Typically, only a single carrier frequency would be disrupted
          due to narrowband interference. The spread spectrum technologies usually
          have a workaround for problems with this type of interference without
          additional administration or configuration.


          The narrowband interference can be identified using a spectrum analyzer
          and disabled from the network. Alternatively, some wireless LAN vendors
          packaged a software spectrum analyzer with the wireless client driver




D-Link Australia & NZ                                                       Page 58 of 85
Version 1.0
                                                                            Wireless LAN


          software. The administrator will be able to know which is the available RF
          that is present in a given area.


      -   All-band Interference
          All-band interference refers to any signal that interferes with the entire RF
          band. Technologies like Bluetooth, which utilize the 2.4 GHz ISM band, has
          a significant interference with the 802.11 RF signals. The other source of all-
          band interference includes microwave oven. A spectrum analyzer can detect
          this kind of problem.


          When all-band interference is present, it is advisable to change to a different
          technology such as 802.11a, which uses the 5 GHz UNII bands. However,
          changing devices can be very costly. The alternative solution is to find out
          the source of the all-band interference and remove it from the space.


      -   Weather
          Severely adverse weather conditions can affect the performance of a
          wireless LAN. The common weather occurrences such as rain, hail, snow
          and fog do not have severe impact on wireless LANs. However, extreme
          occurrences of wind, fog and smog can cause degradation or even
          downtime of your wireless LAN.


          Wind does not affect radio waves or RF signal, but can affect the positioning
          and mounting of outdoor antennas. A strong wind can easily move the
          antennas and cause a complete degradation of signal between two
          antennas. It is important to secure the antennas and cables in locations
          where hurricanes or tornadoes occur frequently.




D-Link Australia & NZ                                                      Page 59 of 85
Version 1.0
                                                                             Wireless LAN




                                                                        Beam
                                                                        arrives at
                                                                        the receiver

                                            No wind




                                                                        Beam
                                                                        misses the
                                                                        receiver
                                          Wind moves
                                          the antenna




          Figure 54. Antenna Wind Loading in a Point-to-Point Network


          When very thick fog or smog settles, the air within this fog becomes very still
          and begins to separate into layers. The fog itself does not cause the
          diffraction of RF signals. The stratification of the air within the fog does.
          When the RF signal goes through these layers, it will bend.


          Lightning can strike the wireless LAN component such as antenna or any
          object nearby. If the wireless LAN devices are not protected by a lightning
          arrestor, the lightning strikes of the nearby object can damage the internal
          components. Lightning can also affects wireless LANs when it charges the
          air through the RF waves after striking an object that is located in between
          the transmitter and receiver.


      -   Adjacent Channel and Co-Channel Interference
          Adjacent channels are channels within the RF bands that are being used
          side-by-side. For example, channel 1 is adjacent to channel 2, which is
          adjacent to channel 3 and so on. These adjacent channels, which are 22
          MHz each, overlap each other with the center frequencies only 5 MHz apart.
          Adjacent channel interference happens when two or more access points


D-Link Australia & NZ                                                        Page 60 of 85
Version 1.0
                                                                          Wireless LAN


         using overlapping channels are located near to each other and their
         coverage cells physically overlap. Adjacent channel interference can
         severely degrade throughput in a wireless LAN.


         It is important to take notice of adjacent channel interference when there are
         co-located access points in the same area. A spectrum analyzer can be
         used to find the problem of adjacent channel interference. It can identify the
         channels that overlap each other.


                                       Adjacent Channel
                                         Interference
                                    Channel 1    Channel 3




         Figure 55. Adjacent Channel Interference


         It is important to move access points on adjacent channels far away from
         each other and make sure that the coverage cells do not overlap.
         Alternatively, use non-over-lapping channels such as channel 1 and 11.


         Co-channel interference will have the same effects as adjacent channel
         interference. It is due to two access points of the same channel overlap
         each other.



                  Co-Channel
                  Interference           Channel 1

                        Channel 1
                                                                          Channel 1




         Figure 56. Co-channel Interference in a Network


         It is best to configure a co-location DSSS system with channel 1 and 11. If it
         is necessary to use channel 1, 6 and 11, the access points should be placed


D-Link Australia & NZ                                                     Page 61 of 85
Version 1.0
                                                                           Wireless LAN


          far apart for minimum interference. This is very common in real-world
          applications.


          Channel reuse is applied in applications where seamless roaming is
          required. It is the side-by-side locating of non-overlapping cells to form a
          mesh of coverage where no cell with the same channel touches each other.


                           Channel 1                        Channel 1

                                          Channel 6




                        Channel 11
                                                              Channel 11


                                           Channel 1




          Figure 57. Channel Reuse


f.    Range Considerations
      When positioning wireless LAN hardware, the communication range of the
      devices must be taken into consideration. There are three things that will affect
      the RF link. They are transmission power, antenna type and location, and
      environment. The maximum communication range of a wireless LAN link is
      where the link begins to become unstable but not totally lost.
      -   Transmission Power
          A higher output power will transmit signal to a greater distance and hence a
          greater range.


      -   Antenna Type
          The antenna can focus the RF energy into a tighter beam to transmit farther.
          When it transmits in all directions, the range of communication is reduced.




D-Link Australia & NZ                                                      Page 62 of 85
Version 1.0
                                                                           Wireless LAN


       -   Environment
           The range of wireless LAN can be decreased in a noisy or unstable
           environment. When the packet error rate is high, the coverage area will be
           reduced.




Site Survey
A site survey is also sometimes known as facilities analysis. It is a map to a
successful implementation of a wireless network. A site survey is very important to
obtain useful information that is very helpful for long term. If there is no proper site
survey, the installed wireless LAN may not work properly even though thousands of
dollars are spent on it.


A site survey is a process with several tasks in which the surveyor can find out the
RF behavior, coverage, interference and hardware location. Its primary objective is to
ensure the wireless LAN clients have continual strong RF signal strength even if they
are on mobile. Wireless clients should not assume they are connected in a wireless
LAN.


Site surveying involves analyzing a site from a RF perspective to discover the kind of
RF coverage needed in order to meet the customer’s requirement. During the site
survey, the surveyor will need to ask questions. These questions allow the surveyor
to gather as much information as possible to make the best recommendation on
hardware, installation and configuration of a wireless LAN. It will also include finding
the best positioning for the hardware. An organized and accurate documentation will
result in a better design and installation process.


A proper site survey provides detailed specifications that will address coverage,
interference sources, equipment placement, power considerations and wiring
requirements. The documentation serves a guide for network design and for
installation of the wireless communication infrastructure.


Without a site survey, there is no knowledge on the customer’s needs, and there will
be areas without RF coverage. It is also not possible to estimate the cost of the
wireless implementation.




D-Link Australia & NZ                                                     Page 63 of 85
Version 1.0
                                                                            Wireless LAN


Preparing for a Site Survey
The planning of a wireless LAN involves collection of information and making
decisions. There are many basic questions that need to be answered before the
actual work of the site survey. These questions are mostly open-ended so that the
surveyor can obtain more information. These are some topics that need to be
touched on before performing a site survey.
a. Facilities Analysis
   The most basic question is ‘what type of facilities is required by the customers?’
   This question has a big impact on the entire site survey. The coverage area,
   number of users, security requirements, bandwidth requirements, and budget are
   to be determined.


b. Existing Networks
   The surveyor needs to find out if there is any existing wired or wireless network in
   place. If there is an existing infrastructure, the contents need to be known. The
   common information about the existing infrastructure include network operating
   systems, number of current users, current wireless LAN protocols and security
   measures, location of the wired LAN connections and naming convention of the
   infrastructure devices.


   It is also necessary to obtain the current detailed network diagram from the
   network administrator. If there is an existing wireless LAN in place, the site survey
   will be more difficult, especially if it is not properly installed. There may be a need
   to disable existing wireless LAN in order to perform the site survey. There may
   also be a need to upgrade existing wired infrastructure to enhance the throughput
   and security of the wireless LAN.


c. Area Usage & Towers
   The surveyor needs to know whether the wireless LAN is meant for indoor,
   outdoor or both. For outdoor set up, it is necessary to find out whether there are
   frequent weather changes, such as hurricanes or tornadoes in the area. There are
   many possible situations and potential obstacles to install and maintain outdoor
   wireless LAN devices. If there are many obstacles such as trees that block the
   direct signal path of the outdoor wireless link, it may be necessary to build a tower
   on top of the building. This will require a structural engineer and permit to install.
   Water-proofing enclosure for the bridges or access points will be required and
   radomes may be considered for protecting outdoor antennas.

D-Link Australia & NZ                                                       Page 64 of 85
Version 1.0
                                                                           Wireless LAN




   Outdoor wireless connections are vulnerable to security attacks. There should be
   documents that show how far the outdoor wireless LAN can be safely extended
   without significant chance of intrusion. It is necessary to check if there are other
   wireless LAN signals nearby so as to ensure there is no interference.


   For indoor wireless LAN implementation, the documents will show the floor layout,
   firewalls, building structure and wiring closets.


d. Purpose & Business Requirements
   The purpose of having a wireless LAN and the business requirement must first be
   considered before conducting a site survey. This information can be obtained via
   interviews with the network users and management of the organization to find out
   what is expected to be done for the wireless LAN and what applications are going
   to be used. If installation is only for a few wireless LAN clients, there is no
   necessity to implement a high-speed 802.11a network in an organization. The
   wrong recommendation can affect the business goals of the organization.


e. Bandwidth & Roaming Requirements
   The bandwidth and roaming requirement will determine what type of wireless LAN
   technology should be implemented. The necessary speed, range and throughput
   per user must be determined so that a site survey can be performed to meet the
   needs of the users. Each department may have different usage and requirement
   of wireless LAN in their area.


   It is necessary to understand the number of users in a given area so that the
   throughput for each user can be calculated. The surveyor will also need to
   determine the technology to use, such as 802.11b or 802.11a.


   There are different types of data transmission. High bandwidth applications such
   as voice or video will require greater throughput. Analyzing and documenting
   these application requirements before the site survey allows the surveyor to make
   more informed decisions when testing the coverage area.


   The figure below shows the survey diagram for different bandwidth requirements.




D-Link Australia & NZ                                                      Page 65 of 85
Version 1.0
                                                                        Wireless LAN




                2 Mbps data rate




                   5 Mbps data rate




   Figure 58. Survey Diagram at Different Data Rate


f. Available Resources
   The surveyor will need to find out from the network manager the project budget
   and the time allocated for the project. He will need to find out whether there are
   administrators who are trained on the wireless networks. The surveyor may
   request for a blueprint of the layout of the building or facility schematics. The
   diagram will show where the walls, network closets, power outlets, and other
   facilities are located.


g. Security Requirements



D-Link Australia & NZ                                                   Page 66 of 85
Version 1.0
                                                                           Wireless LAN


   In some scenarios, data security is very important. It is necessary to explain the
   pros and cons of different wireless security methods. The surveyor needs to find
   out what the existing security policies are and how to incorporate wireless LAN
   into it without violating the rules.




Site Survey Equipment
Different wireless LAN equipment and tools are required for a site survey. For simple
indoor cases, at least one access point, a variety of antennas, cables and connectors,
a laptop computer with wireless card, and site survey utility software will be needed.


The access point used during the site survey should have variable output power and
external antenna connectors. This variable output power allows easy sizing of
coverage cells during the site survey.


Most of the wireless modules or PC cards will be packaged with site survey utility
software. Most of them offer a link speed indicator and signal strength meter. They
provide general indications of coverage. To perform a quality site survey, signal
strength, noise floor, signal-to-noise (SNR) ratio and link speed are recorded. With
quality site surveying software, site survey measurements can be efficiently
completed with accuracy.


While walking around the intended coverage area, attention needs to be taken
particularly to the SNR measurement. This is because this measurement shows the
strength of the RF signal versus the background noise. This is a good indicator of
whether the wireless client is connected or not. There is no hard and fast rule for this
measurement. In general, SNR measurement of 22 dB or more is a viable RF link.


The signal strength indicator is useful to find out whether there is an obstacle
blocking the RF signal or whether the access point has enough power. The SNR
measurement allows the surveyor to know if the link is clear for considerable viability.
The noise level is for determining RF interference that is causing problem to the link.
All these three measurements are useful in designing and troubleshooting a wireless
LAN.


If the wireless card is able to change the power output during the site survey, it
allows the surveyor to test for near/far or hidden node problems. The link speed

D-Link Australia & NZ                                                     Page 67 of 85
Version 1.0
                                                                            Wireless LAN


monitor utility software is used to measure the wireless link speed. It is useful in
determining the size or shape of the cells at a certain required throughput.


The laptop computer is normally used by the site surveyor for checking the signal
strength and coverage while roaming around the facility. Some surveyors use PDA
due to its battery life and physical size. It is much lighter than laptop computer.
Simple screen-capture software can be beneficial because the screenshots can be
saved and presented as part of the site survey report. The surveyor will make hard
copy documentation of all the findings and hence a lot of paper will be needed.


Outdoor site surveys are generally taking more time, effort and equipment than
indoor site surveys. They are more complex and involved calculations and
configuration. Equipment such as antennas, amplifiers, connectors and cables will be
needed. With the knowledge of the characteristics of the wireless link such as
distance, link speed, and power output in advanced, the type of antenna to use can
then be determined. A pair of walkie talkies for communication between two persons
at different end of the wireless link will make the outdoor site surveys more efficient.


A spectrum analyzer is used to determine if there are any other sources of
background interference, such as narrowband interference.


After using a spectrum analysis, a protocol analyzer can be used to find other
wireless LANs that are present in the same area. It can pick up any packets
transmitted by the nearby wireless LANs and provide detailed information on the
channels in use, distance, signal strength, etc.




Conducting a Site Survey
While conducting a RF site survey, 10% of the time is on surveying and 90% on
walking. A pair of comfortable shoes is needed when performing the site surveys.
The site survey is normally conducted with general tasks of recording non-RF related
information first.


For indoor surveys, most of the information is located and recorded on a copy of the
facilities blueprint or drawing. This information will include AC power outlets,
grounding points, and wired network connectivity points. Equipment needed for
conducting the survey will include ladders for mounting access points. Things to be

D-Link Australia & NZ                                                       Page 68 of 85
Version 1.0
                                                                          Wireless LAN


taken notice of include potential RF obstructions such as fire doors, metal blinds,
metal mesh windows, and the potential RF interference sources such as microwave
ovens, elevator motors, and 2.4 GHz cordless phones.


For outdoor surveys, a lot more paper will be needed. The survey will include a
sketch on the obstructions such as trees, buildings, lakes, etc between link sites, and
the visual and RF line of sight between transmitter and receiver. The calculation on
the link distance will also be needed. Weather hazards such as wind, rain, snow and
lightning need to be taken into consideration.


The next task is gathering and recording data on the RF coverage patterns, coverage
gaps, data rate capabilities, and other RF-criteria.
a.    Range & Coverage Patterns
      It starts by placing an access point in a logical location. This location may not
      be the final location. This access point may be moved many times before the
      proper location is found. Generally, the starting point is the center of the area
      when using omni-directional antennas. When using semi-directional antennas,
      it will start from one end of the intended coverage area. However, it does not
      matter where the starting point is. More importantly, the surveyor will need to
      walk slowly with the laptop, wireless module and site survey utility software
      running. While walking, the surveyor will record data rates, signal strength,
      noise floor and signal-to-noise ration (SNR) for every area in the room. Walking
      too fast may cause dead spots or potential interference to be missed.




                                                  New Coverage Area


                          Initial Coverage Area


      Figure 59. Access Point Coverage Testing




D-Link Australia & NZ                                                     Page 69 of 85
Version 1.0
                                                                            Wireless LAN


      When the best locations for the access points are determined, mark the
      location with a bright-colored and easily removable tape. It is optional to take a
      digital picture on the location as part of the site survey report. Make sure that
      the orientation of the antennas is also taken note of.


      For outdoor coverage areas, the survey is expected to cover farther walking
      distance and include more records. In outdoor implementation of access points,
      there are a limited number of places where you can mount the access points.
      Therefore, there are lesser time for moving access points. There are potentially
      much more interference or blockage to a wireless LAN signal outdoors than
      indoors.


b.    Data Rate Boundaries
      It is necessary to record the data rate boundaries or sometimes known as
      concentric zones around the access points. For example, for 802.11b, the
      recorded data rates will be decreased from 11 Mbps to 5.5 Mbps to 2 Mbps to 1
      Mbps. These boundaries will have the slower data rate areas further from the
      access point than the higher data rates.




                                                                            11 Mbps

                                                  1 Mbps       2 Mbps   5.5 Mbps


      Figure 60. Data Rate Boundaries


c.    Documentation
      When the copy of the facility blueprint is marked with circles, dead spots, data
      rates, and signal strength measurements in key spots, another location will be
      selected and the whole process will be repeated. There will be multiple copies

D-Link Australia & NZ                                                      Page 70 of 85
Version 1.0
                                                                           Wireless LAN


      of the facility blueprint or floor plan after completing all the locations. A
      summary will be created indicating the range and coverage of the access point
      from various locations. The results will then be compared to choose the best
      possible location. Site surveying is a very time-consuming task.


d.    Throughput Tests & Capacity Planning
      Another measurement that can be performed by the site surveyor is to test
      throughput from the various points. The coverage and data rate documentation
      will reflect the user’s experience on the wireless LAN. Throughput test such as
      file transfers to and from an FTP server will provide a thorough view of what the
      user may experience. The test is normally performed with the existing wired
      infrastructure connectivity.


      Planning for the user capacity is very important. The network administrator will
      need to provide the number of users and application using the wireless LAN in
      a given area.


e.    Interference Sources
      The site surveyor will need to determine any existing wireless LANs in use
      within or around the facility. It is best to disable existing radios. The surveyor
      may conduct the site survey outside operating hours.


      The site surveyor needs to know whether there is any potential wireless
      implementation in the near future. This will affect the current implementation
      and the site survey that is being performed.


      Other sources of interference include microwave ovens, 2.4 GHz cordless
      phones or radiology. Such potential interference sources need to be
      documented in the survey report. There may be a need to move or replace
      these devices.


      The surveyor needs to take note of the signal loss from some of the common
      obstructions. The table below shows the approximate signal loss.

                                               Additional Loss
                   Obstruction                                       Effective Range
                                                     (dB)

       Open Space                                      0                  100%


D-Link Australia & NZ                                                     Page 71 of 85
Version 1.0
                                                                            Wireless LAN



       Window (non-metallic tint)                      3                     70

       Window (metallic tint)                         5-8                    50

       Light wall (dry wall)                          5-8                    50

       Medium wall (wood)                              10                    30

       Heavy wall (6” solid core)                    15-20                   15

       Very heavy wall (12” solid core)              20-25                   10

       Floor / Ceiling (solid core)                  15-20                   15

       Floor / Ceiling (heavy solid core)            20-25                   10


      Record the interference source, its location and its effect and potential effect on
      wireless LAN coverage, range, and throughput. Taking pictures of the
      interference sources that are permanent, such as lakes and buildings, will
      serve as a visual reference to the client. Pictures of the potential sources of
      interference, such as young tree or future building sites, will help the client to
      make decision in future.


f.    Wired Data Connectivity & AC Power Requirements
      Some of the best positions are constrained to where the AC power sources and
      the network connectivity exist. If the preferred access point locations have very
      good and valid reason, the client may consider installing new AC power
      sources and new network connectivity point. The client may choose to use
      Power-over-Ethernet (PoE).


g.    Outdoor Antenna Placement
      It is necessary to record the outdoor antenna placement, location and
      availability of potential mounting and grounding points. The lightning arrestors
      used by outdoor antennas require proper grounding. Therefore the antennas
      need to be mounted on special mounting materials.


h.    Spot Checks
      After the wireless LAN is installed, it may not work exactly as it has been
      planned. Spot-checking by the site surveyor after the installation is completed
      is helpful to avoid troubleshooting after the actual implementation. Items that
      should be checked include coverage in perimeter areas, overlapping coverage


D-Link Australia & NZ                                                      Page 72 of 85
Version 1.0
                                                                            Wireless LAN


      for seamless roaming and co-channel or adjacent channel interference in all
      areas.




Site Survey Reporting
After studying the client’s facility thoroughly, all the data necessary to prepare a
proper report for the client will be available. This report will be the map for
implementing the wireless LAN and for future documentation reference by the
network administrator. Re-visiting the site may be necessary to gather more data or
to confirm the initial findings.


The following are the main sections of the documentation in the site survey for the
clients.
a.    Purpose and Business Requirement
      This site survey will include all the contact information for the site survey
      company and the client company. Both companies will have a copy of the
      report.


      Restating the customer’s requirements and providing details on how these
      wireless LAN requirements can be met will show the client what types of
      coverage and wireless connectivity they have requested. The report will also
      include an application analysis that the surveyor has tested with the client’s
      application to assure that the proper implementation of the new wireless LAN
      will provide appropriate coverage and connectivity for the wireless clients.


b.    Methodology
      It will include all the methodology that is used for conducting the site survey. It
      has information on what was done, how it was done, and why it was done.


c.    RF Coverage
      Detailed RF coverage patterns and ranges that specify the requirement are
      reported. The concentric circle drawing on the floor plan or blueprint will also be
      included.


d.    Throughput




D-Link Australia & NZ                                                      Page 73 of 85
Version 1.0
                                                                              Wireless LAN


      Detailed bandwidth and throughput findings that show the location where it is
      likely to be greatest or least using drawing on the blueprint will be included.
      Screenshots of the actual numeric measurements are also recorded.


e.    Interference
      Detailed RF interference and obstruction findings are reported. It will include
      pictures about each source of interference. There are suggestions for removing
      RF interference sources and explanation on how the RF interference sources
      will affect the wireless LAN.


f.    Problem Areas
      Discuss with possible solutions to the RF problems found and documented. It
      will include the recommendations of technologies and equipment that can best
      serve the customer’s needs.


g.    Drawings
      Provide Visio, CAD or other types of drawings and graphical illustrations of how
      the network should be configured. All the site survey findings can be
      documented using words and pictures.


h.    Hardware Placement & Configuration Information
      This section will include the name of each device, the physical location of each
      access point and bridge, the mounting method, the channels used and the
      output power that each access point delivers.


i.    Additional Reporting
      Examples of additional information are interference findings, equipment types
      needed, equipment placement suggestions, etc. Other suggestions such as
      security solutions may be added as an optional service, which is normally
      charged separately.




Wired Equivalent Privacy (WEP)
Wired Equivalent Privacy (WEP) is the only method for security during the early
years for IEEE 802.11 wireless LAN. As its popularity increased, it attracted the
attention of cryptographic community to detect cracks in WEP. Although WEP is not
a complete security solution, it is nonetheless better than no security at all. It can still

D-Link Australia & NZ                                                         Page 74 of 85
Version 1.0
                                                                             Wireless LAN


serve as a barrier for some attacks and especially unprotected networks. Most
attacks require a large amount of transmitted data and for home users, the number of
packets sent is relatively small, therefore, WEP still provides a fairly safe option.


WEP has been designed with the intention of making it difficult to break in. The
objectives of WEP is stated by IEEE 802.11.
-     It has to be reasonably strong: The security relies on the difficulty of
      discovering the secret key through a brute-force attack. It is related to the
      length of the secret key and the frequency of changing the keys. WEP allows
      the changing of key and the frequent changing of the Initialization Vector (IV).


-     It has to be self-synchronizing: WEP is self-synchronized for each message. It
      is important because the data is assumed to be delivered. Each packet is
      separately encrypted.


-     It has to be efficient: The WEP algorithm has to be efficient and may be
      implemented in either hardware or software.


-     It may be exportable: The product is designed with WEP and is able to be
      exported to other countries.


-     It is optional: It is the user’s choice to use WEP.


IEEE 802.11 (1999) defined two levels of security: open and shared key. Open
security means no security. As for shared key, it means that both ends of the
wireless link must know the matching key value. The key is a shared secret between
the trusted parties.




Authentication Phase in WEP
When a wireless station wants to join an access point, it must prove its identification
first. The phase is known as authentication.


The purpose of authentication is for each party to prove that he is who he claims to
be. To the access point, if a device can prove that it is trusted, the device’s MAC
address is true. Hence, it will let this device to join. However, in WEP, no secret
token is exchanged upon authentication, so there is no way to know whether

D-Link Australia & NZ                                                        Page 75 of 85
Version 1.0
                                                                             Wireless LAN


subsequent messages coming from the trusted device may be from an impostor.
WEP has been dropped from the Wi-Fi specification despite it is still in the IEEE
802.11 standard.


For open authentication, the mobile device sends one message requesting
authentication and the access point replies with a success message. For shared
authentication, four messages are exchanged. The mobile device requests
authentication, then the access point sends a challenge message. The mobile device
responds to the challenge with the secret key to prove its identification. If the proof is
accepted successfully, the access point will then send the message.




                                Authentication (Request)


                                Authentication (Success)


                                Open Authentication




                                Authentication (Request)

                               Authentication (Challenge)

                               Authentication (Response)

                                Authentication (Success)

                               Shared Authentication


Figure 61. WEP Authentication


If the access point is operating in open mode, it will always accept the authentication
request and responds with an authentication success message. Proprietary
screening methods with MAC address lists are provided in most access points. The
authentication is refused unless the mobile device’s MAC address is found in the list.
This, however, does not protect against MAC address forgery. It only provides basic
protection against very simple attacks using an off-the shelf Wi-Fi card or against
accidental connection to another person’s network.


D-Link Australia & NZ                                                       Page 76 of 85
Version 1.0
                                                                           Wireless LAN




Shared key authentication is used to prove to an access point that the mobile device
knows the secret key. When the mobile device requests authentication, the access
point will send a challenge text. The mobile device will encrypt this challenge text
with the secret key using WEP and sends it back to the access point. The access
point will check whether the result is encrypted with the correct key. This, however,
does not prove to the mobile device that the access point knows the key.


The benefit of authentication exchange is that it prevents mobile stations to join the
network unless they know the WEP key.




Encryption Phase in WEP
In a network with wireless LAN, data from the operating system or driver needs to
pass to the IEEE 802.11 MAC service layer. A packet of data called MAC service
Data unit (MSDU) arrives at the wireless LAN with instructions to send out. This
MSDU will eventually pop out of the MAC service layer on the destination device and
pass to the operating system or driver for delivery to the target application. Before it
is transmitted, the MSDU is broken into smaller pieces and this process is called
fragmentation. Each fragment is processed for WEP encryption. A MAC header is
added to the front and a checksum is added to the end.


The encryption process treats the data as a block of unformatted bytes. The first step
of encryption is to add some bytes called the Integrity Check Value (ICV). ICV is to
prevent anyone from tampering with the message in transit. In both encrypted and
decrypted messages, a check is made to detect whether any of the bits has been
corrupted during transmission. All the bytes in the messages are combined in a result
called the Cyclic Redundancy Check (CRC). This 4-byte value is added to the end of
the message before transmission. If one bit in the message is corrupted, the
receiving device will show that the CRC value does not match and will reject the
message. CRC will detect accidental errors but it does not provide protection against
intentional errors. This is because an attacker can simply re-compute the CRC value
after altering the message, which will ensure that it matches again.


ICV is similar to CRC except that it is computed and added on before encryption. The
conventional CRC is added after encryption. Since ICV is encrypted, no attacker can
re-compute it when he attempts to modify the message. ICV is computed by

D-Link Australia & NZ                                                      Page 77 of 85
Version 1.0
                                                                          Wireless LAN


combining all the data bytes to create a four-byte checksum. It is then added to the
end.




                             Compute Checksum



                                   DATA                   ICV



                                       Encrypt

Figure 62. Adding the ICV


After the ICV is appended, the frame is ready for encryption. The system will select
an IV value first and append it to the secret WEP key. It will then initiate the RC4
encryption engine. Finally, it will pass each byte from the combined data and ICV
block into the encryption engine. For each byte going in, there is an encrypted byte
coming out until all the bytes are processed. This is called Stream Cipher.


In order for the receiver to know how to decrypt the message, the key number and IV
value must be placed in front of the message.


                    Not Encrypted


                        IV    Key ID             DATA & ICV



                                                   Encrypt

Figure 63. Adding the IV and KeyID bits


The MAC header is attached and the CRC value placed at the end to detect
transmission errors.


In the receive process, the receiver notes that the WEP bit is set and it needs to read
and store the IV value. It will read the key ID to select the correct WEP key. It will
append the IV value and initialize the RC4 encryption engine. There is no difference
between the encryption and decryption processes. After the encryption engine is

D-Link Australia & NZ                                                     Page 78 of 85
Version 1.0
                                                                            Wireless LAN


initialized, the data is run through one byte at a time to reveal the original message. It
then computes the ICV and verifies that the value matches with the received
message before the data portion is passed on for further processing.


==Reference Only==
RC4 Encryption Algorithm
An encryption algorithm is a set of operations that is applied to plaintext to generate
ciphertext. It is only helpful when there is also a corresponding decryption algorithm.
RC4 is the encryption algorithm used by WEP. For RC4, the same algorithm is used
for both encryption and decryption. The strength of an algorithm is measured by how
difficult it is to crack the ciphertext. RC4 is simple to implement and is considered
quite strong if it is used in the right way.


The basic idea behind RC4 encryption is to generate a pseudorandom sequence of
bytes called the key stream that is then combined with the data using an exclusive
OR (XOR) operation.


RC4 has the following properties:
For encryption: Plaintext ⊕ Random = Ciphertext
For decryption: Ciphertext ⊕ Random = Plaintext


To the attacker, it looks random. But to both ends of the link, they can generate the
same random value for each byte processed. It is called pseudorandom. You can
calculate the next byte in the sequence only if you know the key used to generate the
stream. If one does not know the key, it will look random. XOR operation will hide the
plaintext values. XOR is an easy operation for a computer. The challenge is to
generate a good pseudorandom number stream. You need one pseudorandom byte
for each byte of message to be encrypted. RC4 generates such a stream.


There are two phases in RC4: key setup and pseudorandom generation. The first
phase, the key setup algorithm, establishes a 256-byte array with a permutation of
the number 0 to 255. That means all the numbers are present in the array but the
order is mixed up. The permutation in the array is called S-box. The next phase in
RC4 is the pseudorandom generation phase. This phase involves more swapping of
bytes in the S-box and generates one pseudorandom byte per iteration called R.




D-Link Australia & NZ                                                       Page 79 of 85
Version 1.0
                                                                          Wireless LAN


To generate the ciphertext, each byte of the plaintext is XORed with a value of R
produced by the RC4 algorithm. The whole process is done using byte length
additions and swaps.
==End of Reference==


Theoretically, RC4 cannot be considered as a complete secure encryption system
because it generates a pseudorandom key stream, which is not truly random.




Wi-Fi Protected Access (WPA)
The insecurity of wireless LAN is a key concern for most organizations. The demand
for wireless LAN is high and encourages organizations to use third-party or
proprietary solutions to secure their wireless network.


The next generation of wireless security after WEP is IEEE 802.11i. It defines a
better wireless network in terms of a robust security network (RSN). WPA is a
security solution that is based on the current capabilities of existing Wi-Fi products
found in the market. WPA uses existing and well-known standards and protocols to
overcome the weaknesses of WEP.


WEP does not provide any access control to the wireless network. WPA overcomes
this problem by specifying mandatory protocols for secure wireless network. The
mandatory protocols are IEEE 802.1x, Extensible Authentication Protocol (EAP) and
Remote Authentication Dial-in User Service (RADIUS).


IEEE 802.1x
The main purpose of 802.1x is to control access when the user joins the corporate
network. There are three main components.
-     Supplicant, the client device who wants to use the network resource
-     Authenticator, it controls the access to the network
-     Authentication server, it contains information regarding the validity and
      authenticity of the user joining the network. It makes decision whether the user
      is allowed to the network. It can be a simple process inside the access point.


Originally, 802.1x is designed to work with wired LAN. It controls the access so that
not everyone who plug into the wired network can use the network resources. In
wireless LAN, each connection request from a client is considered at an access to

D-Link Australia & NZ                                                     Page 80 of 85
Version 1.0
                                                                            Wireless LAN


the LAN with an invisible wire. The authenticator is the access point that performs
access control when each connection request is treated as an unauthenticated
connection until further approval by the authentication server.




                                                           Authenticator
       Wireless
       Device




                                                                           Network
                                                                           Access
                                             RADIO

                                                     MAC
       Wireless
       Device
                                                           Authenticator

                                                                  Access Point


Figure 64. Logical IEEE 802.1x Ports in an Access Point


Extensible Authentication Protocol (EAP)
EAP was originally designed for point-to-point protocol (PPP). It is used for
establishing and finalizing the authentication process. EAP can carry out different
authentication protocols such as Transport Layer Security (TLS) and Tunnel
Transport Layer Security (TTLS). The benefit of EAP is that it does not depend on a
specific authentication scheme and can be easily used to encapsulate any other
authentication methods.


EAP specifies four types of messages that can be used for communication purpose.
-     Request: messages from the access point to the wireless clients
-     Response: messages from the clients to the access point
-     Success: message from the access point when the network access is granted
-     Failure: message from the access point when the network access is denied.




D-Link Australia & NZ                                                       Page 81 of 85
Version 1.0
                                                                                Wireless LAN



                            Start
                      Request Identity
                      Response Identity
                                                            Response Identity
                                                               Request 1
                         Request 1




                                                                                Authentication Server
                        Response 1




                                            Authenticator
         Supplicant




                                                              Response 1



                                                               Request n
                         Request n
                        Response n
                                                              Response n
                                                                Success
                          Success



Figure 65. EAP Message Flow


The authenticator will first respond with an EAP-Request-Identity message. The
supplicant must respond with an EAP-Response-Identity message. Having obtained
the identity of the supplicant, the authenticator needs to contact the authentication
server to find out whether the supplicant is to be allowed in. The authentication
server will verify that the supplicant really corresponds to the identity it has given.


During the authentication process, the authenticator takes a quick look at each EAP
message that is passed between the supplicant and authentication server. It must
wait until the authentication server indicates whether the supplicant has been
accepted or rejected.




EAP over LAN (EAPOL)
The specification of EAP does not specify how EAP messages are transported from
one place to another within the network. Therefore, the IEEE 802.1x defines an EAP
over LAN protocol called EAPOL. EAPOL provides description on how the EAPOL
can be transported over Ethernet (IEEE 802.3).




D-Link Australia & NZ                                                           Page 82 of 85
Version 1.0
                                                                            Wireless LAN


EAPOL has five types of messages.
-     Start: This is used to search and initiate the authentication process. Start
      messages are sent to the multicast MAC address to see if there is any respond
      from the access point.
-     Key: This is used by the access point to send the encryption keys to the client
      once the client has obtain the authorization to access the network.
-     Packet: EAP messages that are going back and forward are encapsulated in
      this type of EAPOL message.
-     Logoff: This is sent to the access point when the client wishes to disconnect
      from the wireless network.
-     Encapsulated-ASF-Alert: This is not used in WPA or RSN.


Encapsulation of EAP message is performed with EAPOL from the user to the
access point and by RADIUS from the access point to the authentication server.


Like EAP, the RADIUS protocol was not originally designed for wireless network. It
was designed for dial-in access. The RADIUS protocol is a set of functionality
compatible with different types of authentication server. The four basic message
types are Request, Challenge, Accept, and Reject.


Security Layers
In the context of wireless LAN security, three layers are clearly identified. They are
Wireless LAN Layer, Access Control Layer and Authentication Layer. Robust
Security Network (RSN) solution can fit into existing security architectures and
leverage on existing standards.


The Wireless LAN Layer is the worker. It deals with raw communications, advertises
capabilities and accepts applications to join the network. It is also responsible for
encrypting and decrypting the data once the security context is established.


The Access Control Layer is the middle manager. It manages the security context. It
will stop any data passing to or from anyone who does not have a current security
context established. When authentication occurs and the security context is
established, the status will change. The access control layer communicates with the
authentication layer to know when to open the security context.




D-Link Australia & NZ                                                       Page 83 of 85
Version 1.0
                                                                        Wireless LAN


The Authentication Layer makes policy decisions and accepts proofs of identity. It
approves the application for someone who wants to join the network. It has ability to
manage the user database. It solves the key management problems of WEP and
makes it easier to integrate wireless LAN with the overall security management
system.




            Authentication             Authentication
               Server                      Client            Authentication
                                                                 Layer



             Corporate                   Operating
              Network                     System


                                                            Access Control
                                                                Layer
            Authenticator                Supplicant
          (Access Control)



           Wireless LAN                 Wireless LAN
                                                             Wireless LAN
                                                                 Layer
          Access Point                Mobile Device




Figure 66. Relationships of Security Layers




Summary of Wireless LAN
Wireless LAN is a flexible data communication system implemented as an extension
or alternative to a wired LAN. It uses electromagnetic waves to transmit and receive
data over the air without physical cabling. Wireless LAN combines data connectivity
with user mobility using simple configuration.


Antennas are used to convert high radio frequency signals on the cable into
propagated waves in the air. It is important for the network managers to understand
the antenna design to correctly design and administrate the network.




D-Link Australia & NZ                                                   Page 84 of 85
Version 1.0
                                                                          Wireless LAN


A site survey is needed to successfully implement a wireless network. It is a process
to discover the type of RF coverage to meet the customer’s requirement and to
ensure that the wireless LAN clients have continual strong RF signals strength when
they are mobile.


It is a challenge to implement a secure wireless LAN. WEP was the only security
method during the early years for wireless LAN. Although it is not a foolproof security
method, it is still better than no security at all. The next generation of wireless
security is WPA. It is designed to overcome the weaknesses of WEP. It involves
IEEE 802.1x, EAP and RADIUS.




D-Link Australia & NZ                                                     Page 85 of 85
Version 1.0

				
DOCUMENT INFO
Shared By:
Categories:
Stats:
views:76
posted:2/23/2011
language:English
pages:85