Point Coordination Function for IEEE 802

Document Sample
Point Coordination Function for IEEE 802 Powered By Docstoc
					         Simulation of Point Coordination Function for IEEE
              802.11 Wireless LAN using Glomosim

            Jeevan Chittamuru, Arunachalam Ramanathan & Manoj Sinha,
                    Dept. of Electrical & Computer Engineering,
                        University of Massachusetts, Amherst.
                   {jchittam, aramanat, msinha} @ ecs.umass.edu



                                I.       Introduction

       The IEEE 802.11 medium access protocol (MAC) supplies the functionality
required to provide a reliable delivery mechanism for data over noisy wireless media.
The MAC sub layer is responsible is responsible for the channel allocation procedures,
frame formatting, error checking, fragmentation and reassembly.
       The MAC architecture can be described as shown in Fig. 1.




                                     Fig.1: MAC Architecture

       The fundamental access method of the IEEE 802.11 MAC is Distributed
Coordination Function (DCF) known as Carrier Sense Multiple Access with Collision
Avoidance (CSMA/CA). All stations must support the DCF. The transmission medium
can operate in the contention mode, requiring all stations to contend for access to the
channel for each packet transmitted. The medium can also alternate between the
contention period (CP) and contention free period (CFP). During the CFP, the medium
usage is controlled by the AP, thereby eliminating the need for stations to contend for
channel access. IEEE 802.11 supports three different types of frames: management, data
and control. The management frames are used for station association and dissociation
with the AP, timing and synchronization etc. Control frames are used for handshaking
during the CP, for positive acknowledgements during the CP, to poll during CFP and end
the CFP.


A.     Distributed Coordination Function (DCF)

        The DCF is the fundamental access mechanism used to support asynchronous
data transfer on a best effort basis. All stations must support DCF. The DCF operates
solely in the ad-hoc network1 and either operates solely or coexists with the PCF in an
infrastructure mode2. Contention services promote fair access to the channel for all
stations.
        The DCF is based on carrier sense multiple access with collision avoidance
(CSMA/CA). CSMA/CD (collision detection) is not used because a station is unable to
listen to the channel for collisions while transmitting. In IEEE 802.11 carrier sensing is
performed at both the air interface, referred to as Physical Carrier sensing, and at the
MAC sub layer, referred to as Virtual Carrier sensing. Physical carrier sensing detects
the presence of other IEEE 802.11 WLAN users by analyzing all detected packets. A
source station performs virtual carrier sensing by sending MPDU duration information in
the header of request to send (RTS), clear to send (CTS) and data frames.

B.     Point Coordination Function (PCF)

        The PCF is an optional capability provided by IEEE 802.11. It provides
contention-free frame (CF) transfer and is only usable on infrastructure network
configurations. This access method uses a point coordinator (PC), which shall operate at
the access point (AP) of the BSS, to determine which STA currently has the right to
transmit. The operation is essentially that of polling, with the PC performing the role of
the polling master. The PCF relies on the PC to perform polling, enabling polled stations
to transmit without contending for the channel. The AP within each BSS performs the
function of PC.

                                 II.    Interframe Space (IFS)

The time interval between frames is called the IFS. A STA shall determine the medium
is idle through the use of the carrier-sense function for the interval specified. The IFS
intervals are mandatory periods of idle time on the transmission medium. Four different
IFSs are defined to provide priority levels for access to the wireless media: short IFS
(SIFS), point coordination function IFS (PIFS), DCF-IFS (DIFS), and extended IFS
(EIFS). Fig 2 shows relationships between different timings.
                                   Fig 2: Interframe Space

A.      Short IFS (SIFS)

        SIFS is the shortest of the interframe spaces. The SIFS is the time from the end
of the last symbol of the previous frame to the beginning of the first symbol of the
preamble of the subsequent frame as seen by the air interface. Using the smallest gap
between transmissions within the frame exchange sequence prevents other STAs, which
are required to wait for the medium to be idle for a longer gap, from attempting to use the
medium, thus giving priority to completion of the frame exchange sequence in progress.

B.     PCF IFS (PIFS)

         The PIFS shall only be used STAs operating under the PCF to gain priority access
to the medium at the start of CFP. A STA using the PCF shall be allowed to transmit
contention free traffic after its carrier sense mechanism determines that the medium is
idle at the T* PIFS slot boundary.

C.     DCF IFS (DIFS)

        The DIFS shall be used by STAs operating under the DCF to transmit data frames
(MPDUs) and management frames (MMPDUs). For the basic access method, when a
station senses the channel is idle, the station waits for DIFS period and samples the
channel again. If the channel is idle, the station transmits an MPDU.

D.     Extended IFS (EIFS)

        The EIFS shall be used by the DCF whenever the PHY (physical layer) has
indicated to the MAC that a frame transmission was begun that did not result in the
correct reception of the complete MAC frame with a correct FCS value. The EIFS is
defined to provide enough time for another STA to acknowledge what was, to this STA,
an incorrectly received frame before this STA commences transmission. Reception of an
error-free frame during the EIFS resynchronizes the STA to the actual busy/idle state of
the medium, so the EIFS is terminated and normal medium access (DIFS) continues
following reception of that frame.
                            III.   DCF Access Procedure

        The CSMA/CA access method is the foundation of the DCF. In general, a STA
may transmit a pending MPDU when it is operating under the DCF access method, either
in the absence of a PC, or in the CP of the PCF access method, when the STA determines
that the medium is idle for greater than or equal to a DIFS period, or an EIFS period if
the immediately preceding medium-busy event was caused by the detection of a frame
that was not received at this STA with a correct MAC FCS value. If, under these
conditions, the medium is determined by the carrier-sense mechanism to be busy when a
STA desires to initiate the initial frame, the random back-off algorithm shall be followed.

                            IV.     PCF Access Procedure

         The PCF provides contention-free frame transfer. The PC resides in the AP. The
PCF is required to coexist with the DCF and logically sits on top of the DCF. The CFP
repetition interval is used to determine the frequency at which the PCF occurs. It is an
option with for a STA to be able to respond to a contention-free poll (CF-Poll) received
from a PC. A STA that is able to respond to CF-Polls is referred to as being CF-Pollable,
and may request to be polled by an active PC. CF-Pollable STAs and the PC do not use
RTS/CTS in the CFP. When polled by the PC, a CF-Pollable STA may transmit only
one MPDU, which can be to any destination (not just to the PC), and may “piggyback”
the acknowledgement of the frame received from the PC using particular data frame
subtypes for this transmission. If the data frame is not acknowledged, the CF-Pollable
STA shall not retransmit the frame unless the PC polls it again, or it decides to retransmit
during the contention period.
         Within a repetition interval, a portion of the time is allotted to contention-free
traffic, and the remainder is provided for contention-based traffic. A beacon frame
initiates the CFP repetition interval, where the AP transmits the beacon frame when the
medium is determined to be idle for one PIFS period. After the initial beacon frame, the
PC shall wait for at least one SIFS period, and then transmit one of the following: a data
frame, a CF-Poll frame, a Data + CF-Poll frame, or a CF-End frame. If there is no traffic
buffered and no polls to send at the PC, a CF-End frame shall be transmitted immediately
after the initial beacon. At the nominal beginning of each CFP repetition interval, all
stations in the BSS update their NAV to the maximum length of the CFP
(CFP_Max_Duration). During the CFP, the only time the stations are permitted to
transmit is in response to a poll from the PC or for transmission of an ACK a SIFS
interval after the receipt of an MPDU. If the PC receives a Data + CF-Ack frame from a
station, the PC can send a Data + CF-Poll + CF-ACK frame to a different station, where
the CF-ACK portion of the frame is used to acknowledge receipt of the previous data
frame. The ability to combine polling and acknowledgement frames with data frames,
transmitted between stations and the PC, has been designed to improve the efficiency.
Fig 3 illustrates the transmission of frames between the PC and a station, and vice versa.




                             Fig 3: PC to Station (STA) transmission

If the PC fails to receive an ACK for a transmitted data frame, the PC waits a PIFS
interval and continues transmitting to the next station in the polling list. Fig 4 illustrates
station-to-station frame transmission during the CFP. The PC may also choose to
transmit a frame to a non CF-Pollable station. After the successful receipt of the frame,
the station would wait for a SIFS interval and reply to the PC with a standard ACK
frame.




                              Fig 4: Station-to-Station transmission
      The PCF controls the frame transfer during CFP. The CFP shall alternate with a
CP, when the DCF controls frame transfers. Fig.5 shows the CFP/CP alternation.




                                 Fig 5: CFP/CP Alternation

        The PC generates CFPs at the contention-free repetition rate, which is defined as
a number of DTIM intervals. The PC controls the length of the CFP, with the maximum
duration specified by the value of the CFP-MaxDuration parameter in the CF Parameter
set at the PC. If the CFP duration is greater than the beacon interval, the PC shall
transmit beacons at the appropriate times during the CFP. The value of the
CFPDurRemaining field shall be zero in beacons sent during the CP. Fig.6 shows a case
where the CFP is two DTIM intervals, the DTIM interval is three beacon intervals, and
the aCFPMaxDuration value is approximately 2.5 times beacon intervals.




                                  Fig 6: Timing relationship

       The PC may terminate any CFP at or before the aCFPMaxDuration, based on
available traffic and the size of the polling list. Because the transmission of any beacon
may be delayed due to a medium busy condition at the nominal beacon transmission
time, a CFP may be foreshortened by the amount of the delay. In the case of a busy
medium due to DCF traffic, the beacon shall be delayed for the time required to complete
the current DCF frame exchange.

B.      CFPMaxduration limit

       The value of CFPMaxDuration shall be limited to allow coexistence between
contention and contention-free traffic. The minimum value for CFPMaxDuration is two
times MaxMPDUTime plus the time required to send the initial beacon frame and the
CF-End frame of the CFP. This may allow sufficient time for the AP to send one data
frame to a STA, while polling that STA, and for the polled STA to respond with a data
frame. The maximum value for CFPMaxDuration is the duration of (BeaconPeriod *
DTIMPeriod * CFPRate) minus [MaxMPDUTime + (2 * aSIFSTime) + (2 * aSlotTime)
+ (8 * ACKSize)], expressed in microseconds. MaxMPDUTime is the time to transmit
the maximum sized MAC frame, expanded by WEP, plus the time to transmit the PHY
preamble, header, trailer, and expansion bits, if any. This allows sending at least one data
frame during the CP.


               V.     Changes Made to Implement PCF over DCF:

A.      Timing Parameters:

        The following timing parameters were added to support PCF timing
        specifications according to IEEE802_11 standard:

(i)     M802_11_PIFS = (M802_11_SIFS + M802_11_SLOT_TIME)
        Time for which the PC node will wait (sensing medium to be idle) at the
        beginning of PCF period

(ii)    M802_11_Min_Data_Transfer (7784*MICRO_SECOND)
        Time required by a PC node to send one data frame to STA and for the polled
        STA to respond with one data frame. This helps PC to determine if it has enough
        time to transfer a data packet during the remaining CFP period. This is also
        referred to as two times Max.MPDUT time.

(iii)   Min. Value of M802_11_CFPMaxDuration (7793*MICRO_SECOND)
        Time taken to transmit a Beacon frame, two times maximum frame size (max.
        MPDU) and a CF_END frame

(iv)    M802_11_CFPDuration (350000*MICRO_SECOND)
        Time taken to support 50 PC polls and corresponding data transfers during a
        single PCF period

(v)     M802_11_CFPRep_Interval (700000*MICRO_SECOND)
        Total time interval or cycle for a single PCF and a single DCF operation

(vi)    M802_11_Beacon_Interval (700000*MICRO_SECOND)
        In our design we assume that the beacon interval and the CFPRep_Interval are the
        same. So, a Beacon frame is transmitted only at the beginning of a PCF (i.e.) CFP
        period.


B.      Glomo MAC 802_11 Structure:

        The following new elements were added to the “glomo_mac_802_11_str” in the
        file 802_11.h to support PCF mode,

(i)     M802_11_MacModes mode  mode = 0 (DCF_Period)
                                     mode = 1 (PCF_Period)
        The above structure element keeps track of the current mode of operation
        (PCF/DCF) for each node.

(ii)    unsigned int timerSequenceNumber_PCF

        To keep track of the timer sequence number, while starting and canceling timers
        in the PCF period.


C.      M802_11_MacStates:

        The following states were added to the M802_11 Finite State Machine to support
        both DCF and PCF:

(i)     M802_11_S_WF_PIFS
        The Point Coordinator (PC ->node 0) goes into this state when there is a packet to
        send from the network layer, the radio Status is IDLE and the NAV value is zero.

(ii)    M802_11_X_BEACON
        The PC node changes to this state when it transmits Beacon Frame to the STA’s

(iii)   M802_11_X_PC_POLL
        When PC transmits a Poll (without any data frame) to a particular node in the
        polling list.

(iv)    M802_11_X_CF_END
        When PC transmits CF_END at the end of PCF period.

(v)     M802_11_X_POLL_ACK
        When the node that receives a Poll from the PC has no unicast data to send from
        its network layer and sends only an ACK for PC’s poll. The same state is used
        when the polled node sends a Broadcast packet and an Ack for PC’s poll.

(vi)    M802_11_X_POLL_ACK_DATA
        When the node that receives a Poll from the PC has data to send to any particular
        node (including PC) and Ack to the PC’s Poll
(vii)   M802_11_X_CF_ACK
        State of the node (other than PC) when it transmits Ack for the data it received
        from another STA

(viii) M802_11_S_WFNAV_PC
       When PC receives a data frame (during PCF), which is not for itself, then it sets
       its NAV value. This prevents the PC from polling the next node till the STA to
       STA transfer is over.

(ix)    M802_11_S_WF_CF_ACK
        State when the node that receives a Poll has sent data to another STA or to the PC
        and waits for ACK from the STA or the PC.

(x)     M802_11_X_PC_BROADCAST
        When PC sends a broadcast data to all the nodes during the PCF period.

(xi)    M802_11_X_PC_POLL_DATA
        When PC transmits a Poll and Data to a particular node in the polling list. This
        happens when the network layer of the PC has data for the next node to be polled.

D.      M802_11_MacFrameType:

(i)      M802_11_BEACON
        A Long Control Frame (20 Bytes) transmitted by the PC node to all the STA’s as
        a broadcast frame. The STA’s extract the NAV value from this frame and start
        their timers for the NAV period.

(ii)    M802_11_CF_END
        A Short Control Frame (14 Bytes) transmitted by the PC node to end the PCF
        period or the Contention Free Period. All the STA’s that receive this frame cancel
        their NAV timers it had not expired.

(iii)    M802_11_CF_END_ACK
        A Short Control Frame transmitted by the PC node to acknowledge the data frame
        it received from a particular STA and to end the PCF period.

(iv)    M802_11_CF_POLL
        A Long Control Frame transmitted by the PC node when it Polls an STA in the
        polling list without sending any data.

(v)     M802_11_CF_POLL_ACK
        A Long Control Frame transmitted by the PC node when it Polls the next STA in
        the polling list and sends an acknowledgement for the data frame it received from
        the previously polled STA.
(vi)    M802_11_CF_POLL_ACK_DATA
        A Data Frame Header (28 Bytes) sent by the PC node when it sends a Poll and
        Data to the next STA and acknowledgement for the data frame it received from
        the previously polled STA.

(vii)   M802_11_CF_POLL_DATA
        A Data Frame Header sent by the PC node when it sends a Poll and Data to the
        next STA in the polling list.

(viii) M802_11_CF_ACK_DATA
       A Data Frame Header sent by the polled STA when it sends a Data frame to any
       other STA or PC and acknowledgement to the Poll.


(ix)    M802_11_CF_ACK
        A Short Control Frame sent by a polled STA when it has no data to send and
        sends only an acknowledgement for the PC’s poll.


(x)     M802_11_ACKtoSTA
        A Short Control Frame transmitted by a node (other than the PC) when it sends an
        acknowledgement for the data it received from the polled STA.

E.      Polling Scheme used for PCF implementation:

        Round Robin algorithm: We have used round robin scheme for polling the poll
aware nodes. In this scheme, the PC maintains a polling list of all the nodes and polls all
the nodes sequentially. The PC polls the next node (at which PCF was relinquished last
time) when the next PCF starts. PC sends data to a node during PCF, only when it has
data to be sent to the node that is going to be polled.
        The above scheme results in a reduction of the throughput, as the PC keeps on
polling nodes in the polling list, even when these nodes don’t have any data to send. To
improve the throughput, we can make a dynamically reconfigurable polling list, wherein
the PC will poll the nodes depending on the history. So, if some of the nodes don’t have
data to send after being polled for a couple of times, they should be dropped from the list.
However, all the pollable nodes will be present when the next PCF period starts.
        The disadvantage of the current implemented scheme is that, the PC sends data to
a node that it is going to poll, only when it has a top packet in its FIFO stored for that
node. This results in substantial reduction in the throughput as can be seen later on in the
plots. The solution to this problem will have to involve a kind of scan function. This
scan function will search a packet (from the FIFO) for the node that is going to be polled.
              VI.      OPERATION SEQUENCE DURING PCF PERIOD:

              PC NODE:

                                         DCF Period




                                Wait For Radio Idle & NAV = 0

                        No
                                        Medium Idle for
                                        PIFS Period?


                          Yes
                                       Transmit Beacon

               B
                              Enough PCF Time to                No
                              send Max. Size Data to                          Pending Ack
                              one STA ?                                       to STA?

                                                                 No                                       Yes
                          Yes
                                       Network                       Tx. CF_END             Tx. CF_END
                                       has packet                    Frame after            + Ack Frame
                                       to send ?                     SIFS                   after SIFS

                No
                                                    Yes

Transmit Pending Ack +                        Unicast           No                 DCF
Poll to next STA in Polling                   Packet?
List after SIFS

                                        Yes


                          Transmit Pending Ack +            Transmit   Broadcast
                          Poll+ Unicast Data to             Packet & Increment
                          next STA in Polling List          Broadcast Packet Tx
                          after SIFS                        Count



                                                                        B
                          Increment Polled
                          Node Counter



                                   A
                                     A



      PIFS Timeout

                       Wait for POLL_ACK from
          B            STA for max. of PIFS
                       Period

                                           Rx. ACK
                                                      Yes
                             Waiting for                       Increment Unicast
                             Data Ack                          Pkt. Tx Count
                             from STA?


              ACK               No
                                                   ACK + STA Data
      B
                         ACK Type?


               ACK + PC Data                                        Start Timer for
                                                                    STA-STA transfer
                               Unicast                              period
                               Data?


              Yes                             No                           B


Increment Unicast                          Increment
Pkt. Rx Count.                             Broadcast Pkt. Rx
Set Pending Ack                            Count.




                         B
STA:


                                               DCF Period


                                                           Start of PCF Period

                                      Complete Pending Data
                                      Transmit. & Receive



                                                   Idle
                       Rx. CF_End                                   Rx.Beacon Frame
Cancel NAV             Frame
Timer &
                                        Start NAV Timer for
Reset NAV
                                        PCF period from
                                        Beacon Frame


                    Timeout                                                              C
   DCF                                  Wait for NAV Timeout

                                                                          Rx. PC Poll.
                 Rx. Broadcast Pkt.
                                                                                               Increment
       C                                                       Poll +Data?                     Unicast Pkt.
                         Increment Broadcast
                                                                                      Yes      Rx. Count
                         Packet Rx. Count
                                                          No
                                         No
           Transmit Ack for                                Network Has
           PC’s Poll                                       Packet to
                                                           Send?


                                                  Yes
                   C
                                                          Unicast
                                                          Packet?
                               Yes                                                            No


                                Tx. PC Ack +                                     Tx. PC Ack +
                                Unicast Data to                                  Broadcast packet
                                PC/STA


               Timeout                                                       Increment Broadcast
       C                        Wait for Ack                                 Pkt. Tx Count.
                                from PC/ STA



                              Increment Unicast                                          C
                              Pkt. Tx. Count
                   VII. IEEE 802.11 PCF/DCF STATISTICS

Global Preset conditions:
1. Beacon repetition interval             - 700 ms
2. CFP repetition interval                - 700 ms
3. CFP duration (50% PCF case)            - 350 ms
4. Channel Bandwidth                      - 2Mbps
5. Channel frequency                      - 2.4 GHz
6. Terrain Range (X & Y)                  - 200 m
7. PC node                                 - node 0
8. Simulation Time                         - 5 minutes


A.     Test Conditions:
       The statistics are obtained by running simulation for a given traffic involving a
variety of sessions such as TELNET, FTP and CBR, though a fixed sets of nodes. The
variation of throughput with the number of total nodes is observed for some
representative traffic involving PC-STA, STA-PC, STA-STA, for100% PCF, 100 %
DCF and 50%DCF+ 50% PCF cases respectively.


       PC was also made to involve in data transmission besides controlling the rest of
the nodes during PCF.
B.     Traffic:
       The traffic used for the simulation is as listed below.
       1. TELNET           0     1 100s    0s
       2. FTP          2        0 500     0s
       3. FTP          2        3 500     0s
       4. FTP          3        4 500     0s
       5. FTP          0        4 500     0s
       6. CBR          3        0 10000   512 0.8s 91.39s 248s
                            VIII. Throughput (bps) for various combinations of PCF and DCF

                                                        100 % PCF case


                                                               100 % PCF

                      400,000
                      350,000
    Throughput(bps)




                                                                                                            2-0 ( STA - PC)
                      300,000
                      250,000                                                                               2-3( STA-STA)

                      200,000
                                                                                                             3-4( STA-STA)
                      150,000
                                                                                                             0-4( PC- STA)
                      100,000
                       50,000
                               0
                                       5       10        20       50        60       80       100
                                                         Number of nodes

                                          Figure 7
         In the 100% PCF case (Fig. 7), as the traffic is only through the nodes 0 to 4, the
throughput is very high when the number of polled nodes is less. With the increase in the
number of polled nodes, the throughput is continuously decreasing as expected. This is
due to the increasing overhead of polling the nodes, which are not contributing to the
traffic.

                                                              100 % PCF
  Number of                                            Throughput (bps) between various nodes
    Nodes
                                    2-0 (STA - PC)         2-3(STA-STA)           3-4(STA-STA)           0-4(PC- STA)

                       5                   168,549              207,230              380,876               198,592
                      10                   114,840              159,015               281,728              143,899
                      20                   77,755               113,511               176,685               91,353

                      50                   45,320               54,835                 81,624               36,290

                      60                   39,140               44,791                 68,931               29,873

                      80                   28,397               31,537                 51,356               22,557

                      100                  19,790               23,296                 39,805               17,806
                              Table 1: Simulation results showing the throughput measured for 100% PCF
                                                    100 % DCF case


                                                            100% DCF

                     350,000
                     300,000
   Throughput(bps)




                     250,000
                                                                                                      2-0 ( STA - PC)
                     200,000                                                                           2-3( STA-STA)
                     150,000                                                                            3-4( STA-STA)
                                                                                                         0-4( PC- STA)
                     100,000
                      50,000
                           0
                                  5       10       20       50      60       80      100
                                                  Number of nodes

                                      Figure 8
       In the 100% DCF case (Fig. 8), the variation in the number of nodes doesn’t have
much effect on the throughput because, only the nodes which contribute to the traffic
contend for the channel. The throughput remains rather constant. However, if the number
of nodes involving in traffic also increase correspondingly, the throughput in DCF case
may be impacted.


                                                          100 % DCF
  Number of                                         Throughput (bps) between various nodes
    Nodes
                                2-0 (STA - PC)          2-3 (STA-STA)         3-4 (STA-STA)            0-4 (PC- STA)

                      5               187,001               153,724               304,608                 280,738

                     10               182,774               152,857               314,307                 275,365

                     20               185,063               149,308               304,468                 285,810

                     50               175,526               147,921               295,679                 267,153

                     60               166,195               134,713               269,650                 238,851

                     80               186,978               142,108               288,847                 249,823

                     100              230,253               127,704               246,489                 163,967
                           Table 2: Simulation results showing the throughput measured for 100% DCF
                                              50% DCF + 50% PCF case


                                                       50%DCF+50%PCF

                     400,000
                     350,000
   Throughput(bps)




                                                                                                           2-0 ( STA - PC)
                     300,000
                     250,000                                                                               2-3( STA-STA)

                     200,000
                                                                                                           3-4( STA-STA)
                     150,000
                                                                                                            0-4( PC- STA)
                     100,000
                      50,000
                            0
                                   5        10       20       50       60       80       100
                                                     Number of nodes

                                                       Figure 9

        The throughput of the system (shown in Fig. 9) reduces (with a greater slope) as
the number of nodes increases initially (for the same reason as described above).
However, the reduction in the throughput is not as much as it was in 100% PCF, because
of the approximately constant throughput contributed by DCF.

                                                  50 % DCF + 50% PCF
  Number of                                         Throughput (bps) between various nodes
    Nodes
                                 2-0 (STA - PC)         2-3(STA-STA)          3-4(STA-STA)           0-4(PC- STA)

                      5                178,917              183,521               346,063                  231,957

                     10                152,065              154,671               300,422                  207,521

                     20                121,809              131,832               244,748                  190,491

                     50                109,157              110,148               193,044                  158,328

                     60                104,580              105,962               185,510                  152,615

                     80                100,641               93,385               174,606                  146,001

                     100               92,679                91,661               162,252                  141,873
                     Table 3: Simulation results showing the throughput measured for 50% PCF and 50% DCF
IX.    Conclusions

         PCF was successfully integrated with the DCF module in IEEE802.11.Various
combinations of PCF and DCF were simulated using Glomosim and the results have been
tabulated and plotted. The following conclusions can be made from the simulation
results:

1.    In PCF mode, as the number of nodes increases, the throughput is found to
      decrease. This is because of the increased overhead in polling the nodes.

2.    If the number of nodes and the traffic between them is high, DCF has poor
      performance as the number of collisions is high. But by dynamically varying the
      polling list based on data transmission history of the nodes, the throughput in PCF
      can be improved.

3.    The combination of PCF and DCF can be used to sustain both synchronous and
      asynchronous traffic. Depending on the type of traffic the percentage of DCF and
      PCF can be varied dynamically.


X.    References:

[1]   IEEE 802.11 standard for Wireless LAN.

[2]   IEEE Communication Magazine, September 1997.

[3]   Reading, IEEE 802.11 Handbook: A Designer’s Companion.

[4]   Kernighan & Ritchie, Programming in C.

[5]   Glomosim and Parsec Manual from UCLA.




                             *** End of Document ***

				
pptfiles pptfiles
About