Wireless LANs

Document Sample
Wireless LANs Powered By Docstoc
					 Analysis of IEEE 802.11e and
Application of Game Models for
Support of Quality-of-Service in
 Coexisting Wireless Networks

        Stefan Mangold
    ComNets Aachen University
         30-June-2003
                          Outline
IEEE 802.11 wireless LAN
   Brief introduction: Distributed Coordination Function (DCF)
IEEE 802.11e QoS extension
   Overview: Enhanced DCF (EDCF)
   Achievable throughput with the EDCF
   Model for achievable throughput per priority
   Result evaluation with WARP2
Overlapping radio networks in unlicensed bands
   Game model of competition
   Result evaluation with YouShi
   Analysis of competition scenario: stability, expected
   outcomes
   Cooperation in repeated games
Conclusions

               Stefan Mangold - ComNets Aachen University        1
       Motivation of this Thesis
IEEE 802.11 is the dominant radio system for wireless
Local Area Networks (LANs):
   Today’s Wireless LANs cannot support Quality of Service
   (QoS)
   However, the demand is growing for new applications with
   QoS requirements
   Wireless LANs operate in unlicensed frequency bands,
   where they have to share radio resources
Problems/Questions:
   How to support QoS in wireless LANs?
   If wireless LANs can support QoS, what level of QoS can be
   maintained in unlicensed frequency bands?


    New methods to support QoS in wireless LANs are
        developed and evaluated in this thesis.

               Stefan Mangold - ComNets Aachen University       2
      IEEE 802.11 Wireless LAN
Radio standard for data transport system that covers
ISO/OSI layer 1 and 2:
   Multiple Physical (PHY) layers: .11/.11a/.11b/.11g
   One common Medium Access Control (MAC) layer
Here: IEEE 802.11a PHY
   OFDM multi-carrier transmission
   Up to 54Mbit/s (@PHY)
5 GHz unlicensed band                                                                                    ne          en
                                                                                                                        t
                                                                              ne                      la           em
                                                                                                    lp           ag ane
   Shared resources                                                       p la
                                                                     er                      t ro              an l
                                                                  us                     c on                 m p

Main Service:                    4
                                     transport layer
                                                             logical link control
                                                                  sublayer                 MLME-

   MSDU Delivery                      network layer
                                                                          MAC-SAP           SAP

                                 3                            medium access
                                                                                          MLME
                                                              control sublayer
                                     data link control

Reference model 
                                           layer                                                                SME
                                 2
                                                              PLCP sublayer
                                                                                          PLME
                                      physical layer
                                                               PMD sublayer
                                 1

                                OSI reference model                                IEEE 802.11

                Stefan Mangold - ComNets Aachen University                                                                  3
                                 Medium Access
Distributed Coordination Function (DCF)
    Listen before talk: CSMA/CA
Binary exponential backoff
    Contention window increases with each retransmission
Received MPDUs (data frames) are acknowledged
                        DIFS                                             with 802.11a:
                                                                          slot:9us     SIFS: 16us     DATA
                       PIFS                                               PIFS: 25us DIFS: 34us

       SIFS           SIFS                                                                           SIFS
                ACK                                                       RTS


                                          Contention Window                          SIFS     CTS
       busy                                (counted in slots,
      channel                       9us per slot, 15 slots in 802.11a)

                  defer access   count down as long as medium is idle,                        time
                                   backoff when medium gets busy



Variable frame body sizes (up to 2312 byte)
One queue per station
Collisions occur if many stations operate in parallel
                          Stefan Mangold - ComNets Aachen University                                         4
 IEEE 802.11 Wireless LAN Basics
MAC protocol is distributed (simple and successful)
    One queue per station (station = backoff entity)
    MSDU can be fragmented into multiple MPDUs
    RTS/CTS helps to improve efficiency
QoS involves achieving a minimum MSDU Delivery
throughput and MSDU Delivery delays not exceeding a
maximum limit
    Delay variation and loss rate are often considered
IEEE 802.11 Task Group E (TGe) defines QoS mechanisms
to be integrated into the legacy 802.11 MAC
    This supplement standard is referred to as IEEE
    802.11e (here: draft 4.0)

            QoS Support in legacy 802.11?  no!


                Stefan Mangold - ComNets Aachen University   5
   802.11e Medium Access: HCF
Contention-based medium access: EDCF
Different EDCF parameters per Access Category (AC)
   DIFSAIFS[AC]                                                       CWmaxCWmax[AC]
   CWminCWmin[AC]                                                     (PF=2PF[AC]*)
                                                            CW[AC=low]
                                                                         aSlotTime
                          AIFS[AC=low]
                                                                                       with 802.11a:
                       AIFS[AC=med.]              low                                   aSlotTime: 9us
                                                            backoff
                                              priority AC                               SIFS: 16us
                       AIFS[AC=high]                                                    PIFS: 25us
                          (=PIFS)                                                       DIFS: 34us
                                          medium            backoff                     AIFSN: 1…10[slots]
                         PIFS            priority AC                                    AIFS: >=PIFS

       SIFS             SIFS                                                  AIFS[AC] =
                 ACK                         high                RTS          SIFS + aSlotTime * AIFSN[AC]
                                          priority AC
                                  earliest channel access
                                    for high priority AC                     SIFS    CTS
        busy
       channel                           CW[AC=high]

       DCF: Random backoff counter is         EDCF: Random backoff counter is
            selected from interval 0...CW.          selected from interval 1...CW+1.       time
            Minimum interframe space is DIFS.       Minimum interframe space is PIFS.
            Earliest channel access is DIFS.        Earliest channel access is DIFS.

                                                                                 *) not in current draft standard

                         Stefan Mangold - ComNets Aachen University                                                 6
         Achievable Throughput
Three different EDCF parameter sets
      AC (priority): higher           medium(=legacy)       lower
        AIFSN[AC]:      2                    2                9
       CWmin[AC]:       7                   15                31
      CWmax[AC]:       1023                     1023        1023
         PF[AC]:       24/16                    32/16       40/16


Question: achievable throughput per backoff entity in an
isolated scenario?  "saturation throughput"
   Isolated scenario means: the same EDCF parameters are
   use by all backoff entities
   Results depend on: frame body length, number of
   contending backoff entities, RTS/CTS, fragmentation
   Approach: WARP2 stochastic simulation and analytical
   model (modifications of Bianchi’s legacy 802.11 model)

               Stefan Mangold - ComNets Aachen University           7
                                                     Legacy (Medium) Priority
                                        512 byte frame body:                                         512 byte frame body, RTS/CTS:
                                    1                                                                                                             1
                                                                     BPSK1/2 (6 Mbit/s)                                                                                               BPSK1/2 (6 Mbit/s)
                                                                     16QAM1/2 (24 Mbit/s)                                                                                             16QAM1/2 (24 Mbit/s)
saturation thrp. (norm.)




                                                                                                      saturation thrp. (norm.)
                             0.8                                     64QAM3/4 (54 Mbit/s)                                          0.8                                                64QAM3/4 (54 Mbit/s)


                             0.6                                                                                                   0.6


                             0.4                                                                                                   0.4

                             0.2                                                                                                   0.2
                                                                                                                                                           with address 4, w/o WEP encrypt.
                                         with address 4, w/o WEP encrypt.
                                    0                                                                                                             0
                                          10   20        40          60           80         100                                                          10    20        40          60            80       100
                                                     number of backoff entities                                                                                       number of backoff entities



                                    2304 byte frame body:                                           2304 byte frame body, RTS/CTS:
                                    1                                                                                                                 1
                                                                      BPSK1/2 (6 Mbit/s)
                                                                      16QAM1/2 (24 Mbit/s)
        saturation thrp. (norm.)




                                                                                                                       saturation thrp. (norm.)
                                   0.8                                64QAM3/4 (54 Mbit/s)                                                        0.8

                                   0.6                                                                                                            0.6


                                   0.4                                                                                                            0.4
                                                                                                                                                                              with address 4, w/o WEP encrypt.

                                   0.2                                                                                                            0.2           BPSK1/2 (6 Mbit/s)
                                                                                                                                                                16QAM1/2 (24 Mbit/s)
                                          with address 4, w/o WEP encrypt.                                                                                      64QAM3/4 (54 Mbit/s)
                                    0                                                                                                                 0
                                          10    20       40          60           80         100                                                           10    20        40          60            80       100
                                                     number of backoff entities                                                                                        number of backoff entities


                                                                          Stefan Mangold - ComNets Aachen University                                                                                                8
                     Low Priority (larger CWmin[AC])
                             512 byte frame body:                                    512 byte frame body, RTS/CTS:
                            1                                                                                      1
                                                      BPSK1/2 (6 Mbit/s)                                                                         BPSK1/2 (6 Mbit/s)
                                                      16QAM1/2 (24 Mbit/s)                                                                       16QAM1/2 (24 Mbit/s)
saturation thrp. (norm.)




                                                                                       saturation thrp. (norm.)
                           0.8                        64QAM3/4 (54 Mbit/s)                                        0.8                            64QAM3/4 (54 Mbit/s)
                                         with address 4, w/o WEP encrypt.
                           0.6                                                                                    0.6
                                                                                                                                  thrp. increases with increasing
                                                                                                                                  number of backoff entities
                           0.4                                                                                    0.4

                           0.2                                                                                    0.2
                                                                                                                                      with address 4, w/o WEP encrypt.
                            0                                                                                      0
                                 10   20            40              60                                                  10       20            40                   60
                                      number of backoff entities                                                                 number of backoff entities



                           2304 byte frame body:                                    2304 byte frame body, RTS/CTS:
                            1                                                                                      1
                                                      BPSK1/2 (6 Mbit/s)                                                            with address 4, w/o WEP encrypt.
                                                      16QAM1/2 (24 Mbit/s)
saturation thrp. (norm.)




                                                                                       saturation thrp. (norm.)
                           0.8                        64QAM3/4 (54 Mbit/s)                                        0.8

                           0.6                                                                                    0.6


                           0.4                                                                                    0.4

                                                                                                                             BPSK1/2 (6 Mbit/s)
                           0.2                                                                                    0.2        16QAM1/2 (24 Mbit/s)
                                                                                                                             64QAM3/4 (54 Mbit/s)
                                            with address 4, w/o WEP encrypt.
                            0                                                                                      0
                                 10   20            40              60                                                  10       20            40                   60
                                      number of backoff entities                                                                 number of backoff entities


                                                          Stefan Mangold - ComNets Aachen University                                                                     9
High Priority (smaller CWmin[AC])
                             512 byte frame body:                                          512 byte frame body, RTS/CTS:
                            1                                                                                            1
                                                            BPSK1/2 (6 Mbit/s)                                                                           BPSK1/2 (6 Mbit/s)
                                                            16QAM1/2 (24 Mbit/s)                                                                         16QAM1/2 (24 Mbit/s)
saturation thrp. (norm.)




                                                                                             saturation thrp. (norm.)
                           0.8                              64QAM3/4 (54 Mbit/s)                                        0.8                              64QAM3/4 (54 Mbit/s)
                                         with address 4, w/o WEP encrypt.
                           0.6                           deviation with higher                                          0.6
                                                                                                                                deviation with higher
                                                         collision probability
                                                                                                                                collision probability
                           0.4                                                                                          0.4

                           0.2                                                                                          0.2
                                                                                                                              with address 4, w/o WEP encrypt.
                            0                                                                                            0
                                    10      20            40                60                                                   10     20            40                60
                                            number of backoff entities                                                                  number of backoff entities



                           2304 byte frame body:                                          2304 byte frame body, RTS/CTS:
                            1                                                                                            1
                                                            BPSK1/2 (6 Mbit/s)
                                                            16QAM1/2 (24 Mbit/s)
saturation thrp. (norm.)




                                                                                             saturation thrp. (norm.)
                           0.8                              64QAM3/4 (54 Mbit/s)                                        0.8

                           0.6                                                                                          0.6


                           0.4                                                                                          0.4   with address 4, w/o WEP encrypt.

                           0.2                                                                                          0.2       BPSK1/2 (6 Mbit/s)
                                                                                                                                  16QAM1/2 (24 Mbit/s)     deviation with higher
                                                                                                                                  64QAM3/4 (54 Mbit/s)     collision probability
                                 with address 4, w/o WEP encrypt.
                            0                                                                                            0
                                    10      20            40                60                                                   10     20            40                60
                                            number of backoff entities                                                                  number of backoff entities


                                                                Stefan Mangold - ComNets Aachen University                                                                         10
            Modified Bianchi Model
                                                           (1-p)/W0
“fire
and                          (1-p)/W0                                                    W0    W0[AC] in EDCF
new“
                       1                  1                               1
        0,0                 0,1                0,2             0,W0-2             0,W0-1

           “coll.“                                                                k:    slot index
                                 p/W1                                             i:    stage index
                                                                                  m:    maximum backoff stage
                                                                                  p:    collision probability
        i-1,0                                                                     W0:   CWmin+1
                                                                                  Wm:   CWmax+1
                                                                        p/Wi
           “coll.“               p/Wi

                       1                  1                                                1
          i,0              i, k=1             i, k=2                           i,Wi-2          i,Wi-1

            “coll.“              p/Wi+1                       m depends on Persistent Factor (PF)
                                                             in the EDCF (proposed), CWmax, and
                p/Wm                                         the retry counter.

                       1                  1                                                             1
        m,0                m,k=1              m,k=2                                        m,Wm-2           m,Wm-1

                                                                           p/Wm
           “coll.“ p/Wm
 stage
index i
                  slot index k                       m   m[AC] in EDCF                  Wm Wm[AC] in EDCF

                            Stefan Mangold - ComNets Aachen University                                               11
              Share of Capacity
Saturation throughput shown so far is only valid for
isolated scenarios
Nice to have, but useless for QoS support:
    For QoS support, a backoff entity needs to know the
    expected throughput in mixed scenarios
    Achievable throughput per backoff entity is referred to as
    "share of capacity"
Question: what is the share of capacity a backoff entity
can achieve in a mixed scenario?
    This is *THE* important question for EDCF QoS support
    Bianchi model does not provide the answer
    There is no solution available until today




                Stefan Mangold - ComNets Aachen University       12
                      Access Probability per Slot
                           AIFSN[higher pr.]=2                                                                         higher pr.
               0.6                                               8 backoff entities per AC[higher]                     medium pr.
                                                                 8 backoff entities per AC[medium]                     lower pr.
               0.5
prob(access)




                                                                 8 backoff entities per AC[lower]
                           AIFSN[medium pr.]=2
               0.4

               0.3                                                       AIFSN[lower pr.]=9
               0.2

               0.1

                0
                     25   34    43     52        61   70   79   88      97     106   115      124    133   142   151     160   169
                                                                     slot [s]




                                                                                                                       higher pr.
               0.6                                               3 backoff entities per AC[higher]                     medium pr.
                                                                 3 backoff entities per AC[medium]                     lower pr.
               0.5         AIFSN[higher pr.]=2
prob(access)




                                                                 3 backoff entities per AC[lower]
               0.4

               0.3         AIFSN[medium pr.]=2

               0.2                                                       AIFSN[lower pr.]=9
               0.1

                0
                     25   34    43     52        61   70   79   88      97     106   115      124    133   142   151     160   169
                                                                     slot [s]


                                         Stefan Mangold - ComNets Aachen University                                                  13
Approximation of Expected Idle Times
 Expected size of contention window
     N[AC] = number of backoff entities of AC
     tau[AC] = probability that a backoff entity is transmitting

                   1                     ! 1    AC N  AC 
  E CW  AC  
               AC                                                       E CW  AC 
                                                                                        
                                         1  1    AC 
                                                                N  AC 

     -persistent CSMA with N                        Bianchi approximation with N
    contending backoff entities per AC
                                                     contending backoff entities per AC



 Access probability per slot
     Expressed by geometric distribution


                                                                                    
                                                                                             N  AC 
                                                                slot  AIFS  AC 
    slot  AC   1  1    AC   1    AC 
                       
                                                                                         
                                                                                         
                                                                                        

                            Stefan Mangold - ComNets Aachen University                                  14
 CSMA Regeneration Cycle Process
                                              C:     inter-AC collision                     H:        high priority access
State transition diagram                      M:     medium priority access                 L:        low priority access

for the Markov chain                       1                                          1                     1                 1

                                          C                                           H                   M                   L

   States C, H, M, L                                                           P1,H           P1,M                 P1,L
                                                    P1,C
   represent busy system                                          1

   States 1, 2, 3...,                                              P1,2
                                                                              P2,H           P2,M                 P2,L
   CWmax+1 represent idle                           P2,C
                                                                  2
   system                                                          P2,3
   Time is progressing in                                          Pslot-2, slot-1          Pslot-1,M            Pslot-2,L
   steps of a slot                                 Pslot-1,C          Pslot-1,H
                                                               slot-1
   State of the chain                                              Pslot-1, slot                                 Pslot,L
   changes with state                              Pslot,C             Pslot,H
                                                                                            Pslot,M

   transition probabilities as                                  slot
                                                                                      PCWmax+1,H
   indicated in the figure                                         Pslot, slot+1
                                                                                                                 PCWmax+1,L
                                                                                            PCWmax+1,M
                                                                   PCWmax, CWmax+1
                                               PCWmax+1,C
                                                               CWmax
                                                                +1                        CWmax = max(CWmax[AC])

                  Stefan Mangold - ComNets Aachen University                                                                      15
                 Markov Chain (1)
Resulting state transition probabilities
    access:      Pslot ,H  slot  High   1  slot  Medium   1  slot  Low  ,
                 Pslot , M  slot  Medium   1  slot  High   1  slot  Low  ,
                 Pslot ,L  slot  Low   1  slot  High   1  slot  Medium  .

    collision:
                 Pslot ,C  slot  High   slot  Medium   1  slot  Low  
                            slot  High   slot  Low   1  slot  Medium  
                            slot  Medium   slot  Low   1  slot  High  
                            slot  High   slot  Medium   slot  Low  .
    idle:                                                     slot  CWmax
                                  
                                                  0,
                 Pslot ,slot 1  
                                  1   Pslot ,H  Pslot , M  Pslot ,L  Pslot ,C  ,
                                                                                         else

                   Stefan Mangold - ComNets Aachen University                                    16
               Markov Chain (2)
Resulting stationary distributions
   high:               CWmax 1       slot 1   
            pH  P1,H   Pslot ,H   Pi ,i 1   p1
                        slot  2        i 1    
                         :   AC High 
                                        
                  (this defines the relative priority of the AC "High")



   other:                 CWmax 1        slot 1   
            p M  P1, M   Pslot , M   Pi ,i 1   p1 :   Medium   p1 ,
                             slot  2       i 1    
                         CWmax 1     slot 1     
            pL  P1,L   Pslot ,L   Pi ,i 1   p1 :   Low   p1 ,
                           slot  2      i 1     
                         CWmax 1     slot 1     
            pC  P1,C   Pslot ,C   Pi ,i 1   p1 .
                          slot  2      i 1      

                 Stefan Mangold - ComNets Aachen University                     17
                                  Result
The priority vector
                                  1
     H ,  M ,  L                    High  ,   Medium  ,   Low 
                                AC 
                             AC




Share of capacity
                                     Thrp  High  H 
                                         sat                 
           Thrpshare  Thrpsat    Thrpsat  Medium   M  .
                                    
                                     Thrp  Low   L       
                                                              
                                         sat                 

    Modified Bianchi model provides the saturation throughput

                    Stefan Mangold - ComNets Aachen University                      18
                                                Scenario & Results (1)
Four backoff entities per AC (4/4/4)
    Variable, legacy and low priority                                                                                                                                 legacy
                                                                                                                                                                      priority
    Results of WARP2 simulation indicate
    accurate approximation
                                                                                                                                                                                        low
                                                                                                                                                                                      priority
                                                                                                                                                     variable
                                                                                                                                                      priority
                                                                                                                                                                 receiving station




                                    0.7        higher priority                                          legacy priority                                               lower priority 

                                                                                                                                           512 bytes frame body, no RTS/CTS
                                    0.6
    share (=thrp.) per AC (norm.)




                                                                                                                                           4+4+4 backoff entities
                                    0.5                                                                    analyt.

                                    0.4                                                                        sim.
                                                                                                                                                          variable pr. (high  low) sim.
                                    0.3                                                                                                                   variable pr. (high  low) apprx.
                                                                                                                                                          legacy pr. sim.
                                                                                                                                                          legacy pr. apprx.
                                    0.2                                                                                                                   lower pr. sim.
                                                                                                                sim.                                      lower pr. apprx.
                                    0.1                                                          analyt.

                                     0

    AIFS:                                 2     2   2    2    2    2    2    2    2    2    2      2       3    4      5    6    7    8    9    9    9      9     9     9    9       9    9      9

   CWmin:                                 7     7   7    7    7    8    9    10   12   13   14     15    15     15     15   15   15   15   15   17   19     21   23    25   27       29   31     31

   16  PF: 24                                 26   28   30   32   32   32   32   32   32   32     32    32     32     32   32   32   32   32   32   32     32   32    32   32       32   32     40


                                                              Stefan Mangold - ComNets Aachen University                                                                                              19
                                                Scenario & Results (2)
10/2/4 backoff entities per AC
    Backoff entities with variable priority                                                                                                                               legacy

    are more dominant, as expected
                                                                                                                                                                          priority



    Results of WARP2 simulation indicate                                                                                                           variable
                                                                                                                                                                                            low
                                                                                                                                                                                          priority
    accurate approximation                                                                                                                          priority
                                                                                                                                                               receiving station




                                    0.7        higher priority                                          legacy priority                                             lower priority 

                                                                                                                                         512 bytes frame body, no RTS/CTS
                                    0.6
    share (=thrp.) per AC (norm.)




                                                                                                                                         10+2+4 backoff entities
                                    0.5
                                                                                  analyt.

                                    0.4                                                sim.                        analyt.
                                                                                                                                                   variable pr. (high  low) sim.
                                    0.3                                                                                      sim.                  variable pr. (high  low) apprx.
                                                                                                                                                   legacy pr. sim.
                                    0.2                                                                                                            legacy pr. apprx.
                                                                                                                                                   lower pr. sim.
                                                                                                                                                   lower pr. apprx.
                                    0.1

                                     0

    AIFS:                                 2     2   2    2    2    2    2    2    2     2     2    2      3   4     5    6      7   8    9    9    9    9       9     9       9      9       9       9

   CWmin:                                 7     7   7    7    7    8    9    10   12    13    14   15    15   15    15   15    15   15   15   17   19   21      23   25      27      29     31       31

   16  PF: 24                                 26   28   30   32   32   32   32   32    32    32   32    32   32    32   32    32   32   32   32   32   32      32   32      32      32     32       40


                                                              Stefan Mangold - ComNets Aachen University                                                                                                  20
                                                Scenario & Results (3)
2/10/4 backoff entities per AC
    Backoff entities with variable priority are                                                                                                                       legacy
                                                                                                                                                                      priority
    more dominant, as expected
    WARP2 simulation results deviate for                                                                                                             variable                               low

    different persistent factors
                                                                                                                                                      priority                            priority
                                                                                                                                                                 receiving station




                                    0.7        higher priority                                             legacy priority                                               lower priority 

                                                                                                                                          512 bytes frame body, no RTS/CTS
                                    0.6
    share (=thrp.) per AC (norm.)




                                                                                                                                          2+10+4 backoff entities
                                    0.5                                                   analyt.

                                    0.4                                                       sim.
                                                                                                                                                    variable pr. (high  low) sim.
                                                                                                                                                    variable pr. (high  low) apprx.
                                    0.3
                                                                                                                                                    legacy pr. sim.
                                                                                                                                                    legacy pr. apprx.
                                    0.2                                                                                                             lower pr. sim.
                                                                                                                                                    lower pr. apprx.
                                    0.1                                               sim.
                                                                                       analyt.
                                     0

    AIFS:                                 2     2   2    2    2    2    2    2    2       2      2    2      3   4    5    6    7    8    9    9     9     9      9       9      9    9       9      9

   CWmin:                                 7     7   7    7    7    8    9    10   12     13      14   15    15   15   15   15   15   15   15   17    19    21     23      25     27   29     31      31

   16  PF: 24                                 26   28   30   32   32   32   32   32     32      32   32    32   32   32   32   32   32   32   32    32    32     32      32     32   32     32      40


                                                               Stefan Mangold - ComNets Aachen University                                                                                                 21
              EDCF Summary
EDCF MAC protocol is distributed (as DCF, simple)
Multiple queues per station (queue = backoff entity)
The presented model can be used for prediction of
expected share of capacity per backoff entity
The model can be extended towards delay and loss
prediction
EDCF supports QoS, but cannot guarantee as resulting
share depends on activity of other backoff entities

        QoS Support in legacy 802.11?  no!
QoS Support in 802.11e EDCF?  yes, but no guarantee!




              Stefan Mangold - ComNets Aachen University   22
   HCF Controlled Medium Access
EDCF cannot guarantee QoS, because of distributed MAC
For guarantee, controlled medium access allows access
right after PIFS, without backoff
Similar to polling in legacy 802.11 (PCF)

                                                                               tolerated by HC 1

                                                optimal CAP allocation     delayed CAP allocation
                                                    time for HC 1              time for HC 1


               time                                            delayed start of TXOP
                                                                                       QoS CF-Poll

          AIFS[AC]              RTS                 DATA (MSDU)


           busy                          CTS                             ACK
          channel
                                                                               PIFS        CAP
                                    EDCF-TXOP gained by contention-based
                                     channel access during contention period            allocation

                                           duration < EDCF-TXOPlimit
             under control ofHC 1




                      Stefan Mangold - ComNets Aachen University                                     23
                                          HCF in Overlapping BSS
Controlled medium access requires an isolated BSS
No other backoff entity must access the medium with
highest priority (after PIFS), otherwise collisions occur!
This is a very strict requirement, and difficult to achieve
in an unlicensed frequency band
Dynamic frequency selection may help, as in HiperLAN/2

                                      512 byte frame body:                                                                  2304 byte frame body:
                             1                                                                                     1
                                               BPSK1/2 (6 Mbit/s)                                                                     BPSK1/2 (6 Mbit/s)
                                               16QAM1/2 (24 Mbit/s)                                                                   16QAM1/2 (24 Mbit/s)
 saturation thrp. (norm.)




                                                                                       saturation thrp. (norm.)
                            0.8                64QAM3/4 (54 Mbit/s)                                               0.8                 64QAM3/4 (54 Mbit/s)

                            0.6             CW=0, AIFSN=1                                                         0.6              CW=0, AIFSN=1


                            0.4                                                                                   0.4

                            0.2                                                                                   0.2


                             0                                                                                     0
                                  1        2                            10                                              1         2                            10
                                        number of HCs allocating CAPs                                                          number of HCs allocating CAPs




                                                          Stefan Mangold - ComNets Aachen University                                                                24
   HCF Controlled Access Summary
 The controlled medium access is often referred to as HCF
 This coordination function is not distributed, it is
 centralized (requires a Hybrid Coordinator)
 It works only in isolated scenarios, which is not a very
 likely scenario in unlicensed bands
 The coexistence problem of overlapping BSSs will be
 discussed in the following

         QoS Support in legacy 802.11?  no!
 QoS Support in 802.11e EDCF?  yes, but no guarantee!
QoS Support with 802.11e HCF?  not in unlicensed bands!




                 Stefan Mangold - ComNets Aachen University   25
Scenario: two BSSs Sharing one Channel
Basic service sets are modeled as players that attempt to
optimize their outcomes
Single stage game: one superframe (~200ms)
Multi stage game: repeated interaction
                                              vectors indicate
                                              "has control over"


                               HiperLAN/2
                                 station


                                                    802.11
                          CCHC                      station
                 802.11 (player 1)
                 station
                                        CCHC
                                      (player 2)
                                                   HiperLAN/2
                                                     station
            CCHC's
            detection ranges

                Stefan Mangold - ComNets Aachen University         26
The Superframe as Single Stage Game
Allocation process during a superframe:
                                                                     nth CCHC superframe = the
                                              SFDUR(n)[ms]               nth single-stage game

              D11(n) [ms]            D21(n) [ms]            DL1(n) = D31(n) [ms]

           d11(n) [ms]       d21(n) [ms]               d31(n) [ms]

                                                                                      allocated
                                                                                     by CCHC1

                                                                                     allocated by
                                                                                        CCHC2
                                                            1...L1 TXOPs allocated
           t11(n)           t21(n)                 t31(n)
                                                            by CCHC1 (here, L1=3)
    TBTT                                                                                      TBTT
             the periodic beacon is successfully
              transmitted by one of the CCHCs                               time


QoS:                                          [ 0...1 ]           
                                QoS           [ 0...1 ]           
                                                                     
                                              [ 0...1 ]           
                                                                     
                         Stefan Mangold - ComNets Aachen University                                  27
     Abstract Representation of QoS
Throughput: normalized share of capacity
                                                       Li ( n )
                                       1
                              i
                             (n)                               d li ( n )
                                    SFDUR( n )          l 1



Delay: normalized resource allocation interval
                                     1
           imax ( n )                      max  Dli ( n ) 
                       ,          SFDUR( n )                l 1...Li ( n ) 1


Jitter: normalized delay variation
                         1
      imax ( n )               max  Dli ( n )  Dli 1 ( n ) 
                      SFDUR( n )                               l 1...Li ( n ) 1

                           Stefan Mangold - ComNets Aachen University                 28
                                   The Player
Player "i" and opponent player "–i" have individual
requirements
Players select demands to meet requirements
Through allocation process, players observe outcomes per
single stage game: observed QoS
                                                              demand of player -i
                                                                                       time
          requirement                                                       i
                                                                                 n
  ireq                                                 n                                      iobs  n
                                                                             dem
                          action ai of player i:  i

                                                                           i  n 
                                                   dem
                           select demand based
  ireq                                          idem  n                                      iobs  n 
                                                                            dem
                   time       on requirement,
                             observation, and            demand                        observation(outcome)
                           estimated demand of
                                  player -i
                                                                    time                               time

                                                                                   allocation process
                                  z -1

This single stage game is repeated with every superframe
Players adapt behaviors in the multi stage game
                          Stefan Mangold - ComNets Aachen University                                           29
    Allocation Process (Formal Description)
    Required:

            dem   dem  
               i        i        obs 
                                    i
                  ,        i  ,                i , i    {1, 2}
            i   i         
            dem   dem      obs 
    If this process can be formally described through an
    accurate approximation, we can analyze
        Expected outcomes (existence of Nash equilibrium (NE))
.
        Stability (convergence to NE)
        Fairness (position of NE in bargaining domain)
    It can be discussed…
        … what QoS support is feasible for the individual players
        (player = CCHC = BSS)
        … what level of QoS can be achieved
        … if mutual cooperation improves the outcome per player

                    Stefan Mangold - ComNets Aachen University                30
                                   Markov Chain
Observed payoffs in a single stage game:
                                                       P23


                             P43          P30          P01           P12
                    p4              p3           p0           p1           p2
                             P34          P03           P10          P21


                                          P41


Stationary distributions:
        p0:     idle channel (EDCF background traffic)
        p1:     player 1 allocates radio resource
        p2:     player 2 backing off while player 1 allocates resource
State transition probabilities:
            2
           dem                                            1      1            1,2
P01                     ,   1,2   0          P12  min 1, dem   dem
                                                                                 dem  0
         2
        dem    1
                              dem                           2 1  2          
                  dem                                      dem        dem      
                             Stefan Mangold - ComNets Aachen University                      31
                 Result and Evaluation
Resulting observations for both players:
                                                  i     i
                          i                      dem dem
                         obs       i     i      i     i
                                    dem dem  dem  dem
      i                  i                             i     i
     obs              dem                         dem  dem  TXOPlimit
              demanded allocation interval       unwanted increase of allocation interval ( delay )

Comparison with simulation results (YouShi):




                      Stefan Mangold - ComNets Aachen University                                      32
                 The Utility Function
Players attempt to meet their requirements
Therefore, players attempt to maximize the observed
payoff (outcome), by using a utility function




                                                                     
               i    i      i      i         i   i      i
       U i  U  ( dem , obs , req )  U ( obs , req ), U i
                  Stefan Mangold - ComNets Aachen University             33
Existence of Nash Equilibrium (NE)
Proposition:    in the Single Stage Game of two
coexisting CCHCs exists a Nash equilibrium in the action
space A.
Proof: show that the outcome (the payoff V) is
continuous in A, and show that it is quasi-concave in Ai.


There exists at least one Nash equilibrium, which can be
calculated as:

                                                                   
                                                            i     
     a  a*  0  grad i V i ( a ) i   , with grad i            
                                                               dem
                                                                   
                                                          
                                                            i     
                                                                     
                                                              dem   

      a=action, V=payoff, N=number of players (N=2)

                  Stefan Mangold - ComNets Aachen University             34
               Pareto Efficiency
Players that take rational actions will automatically adjust
into a NE (because there is at least one NE)
If the NE is unique, the respective action profile can be
predicted as expected point of operation


However, there may exist action profiles in the single
stage game that lead to higher payoffs
If such profiles do not exist, the NE is referred to as
Pareto efficient (Pareto optimal)


Pareto efficiency can be determined by numerical search


Can be shown in bargaining domain … (next page)


                Stefan Mangold - ComNets Aachen University     35
                                              Bargaining Domain
                                     1



                                    0.9


                                                                                          Pareto boundary
                                    0.8
                                                                      Nash equilibrium    

                                    0.7
payoff of player -i: V -i(ai,a-i)




                                    0.6



                                    0.5



                                    0.4



                                    0.3



                                    0.2



                                    0.1       fair share


                                     0
                                          0     0.1    0.2   0.3       0.4     0.5     0.6            0.7   0.8   0.9   1
                                                                                        i    i   -i
                                                                   payoff of player i: V (a ,a )

                                                 Stefan Mangold - ComNets Aachen University                                 36
                                                Strategy: Persist

        1                                                 required                        1
pl1             simulated
                (solid line)                              observed               pl2            simulated
                                                                                                                                          required
                                                                                                                                          observed
                                                          demanded                              (solid line)                              demanded
       0.6                                                                               0.6
                                                                                                                                                      
  1




                                                                                   2
              analyt. apprx.        some variations   arrow indicates that
              (dashed line          because of EDCF   players generally
               not visible)                           attempt to maximize                          analyt. apprx.
        0                                             throughput                          0        (dashed line)

       0.1    simulated        analyt. apprx.          required                          0.1                                       required
                                                                                                simulated
              (solid line)     (dashed line)           observed max                             (solid line)    analyt. apprx.     observed max
                                                       demanded                                                 (dashed line)      demanded

                                                                                                                                                      
 1




                                                                                  2
      0.04
                                                  arrow indicates that
                                                  players generally                    0.023
                                                  attempt to minimize
        0                                         delays                                  0
        0.2     1      1.8       2.6 3.4 4.2 5 5.8            6.6   7.4                   0.2      1      1.8     2.6 3.4 4.2 5 5.8        6.6   7.4
                               time (s), SFDUR = 200ms                                                          time (s), SFDUR = 200ms



 Persist: demand=requirement
 Shown are YouShi simulation results and analytical apprx.
 Poor delay performance for pl.2
                                                Stefan Mangold - ComNets Aachen University                                                                 37
Persist/Best Response/Cooperation
 defect                                                                               1
                                                              pl 1
                                                              pl 2




                                                                     Utilities U1,2
                                                                                                in total, player 1 observes a




                                                                                
persist                                                                                                                                        pl 1
                                                                                                higher payoff than player 2 when
                                                                                                both demand their requirements
                                                                                                                                               pl 2
               both players demand                                               0.4
  coop         requirements throughout
               all stages (BEH-P)                                                0.2
  best                                                                                0
     0.2   1   1.8     2.6 3.4 4.2 5 5.8               6.6    7.4                     0.2   1   1.8     2.6 3.4 4.2 5 5.8               6.6    7.4
defect                                                       pl 1                     1                                                       pl 1
                     time (s), SFDUR = 200ms                                                          time (s), SFDUR = 200ms
                                                             pl 2                                                                             pl 2
                                                                                                           in total, player 2 gains and player 1




                                                                     Utilities U1,2
                                                                                
persist                after 2s, both players change                                                       suffers from playing the best responses
                       their behavior from BEH-P to
                       BEH-B independently, then                                 0.4
 coop                  attempting to improve their                                                neither player is able to improve its outcome
                       individual payoffs                                        0.2              by unilaterally changing its behavior from what
                                                                                                  is demanded after the process converged into NE
  best                                                                                0
    0.2    1   1.8     2.6 3.4 4.2 5 5.8               6.6    7.4                     0.2   1   1.8     2.6 3.4 4.2 5 5.8               6.6    7.4
defect               time (s), SFDUR = 200ms                 pl 1                     1
                                                                                                      time (s), SFDUR = 200ms
                                                             pl 2
                                                                                                       in total, payoffs are higher in cooperation

                                                                     Utilities U1,2
                                                                                
persist                                                                                                than in NE, therefor the NE is not Pareto
                                                                                                       efficient in this example
                                                                                 0.4
 coop
                     both players cooperate after 2s                             0.2
                                                                                                                                              pl 1
  best                                                                                0                                                       pl 2
     0.2   1   1.8     2.6 3.4 4.2 5 5.8               6.6   7.4                      0.2   1   1.8     2.6 3.4 4.2 5 5.8               6.6    7.4
                     time (s), SFDUR = 200ms                                                          time (s), SFDUR = 200ms


                                     Stefan Mangold - ComNets Aachen University                                                                       38
       How to establish Cooperation
Cooperation can be beneficial for both players, and is
established in repeated interactions (multi stage game)
Cooperation and punishment:


                otherwise

                                                      for a number of stages,
       COOPERATE:              PUNISH(1):          depending on discouning factor     PUNISH(n’):
n=n0     BEH-C                   BEH-D                                                  BEH-D

                    opponent                     any behavior          any behavior
                     defects                     of opponent           of opponent




Payoff discounting in multi stage game:
                                            
                                i
                               VMSG        nV i(n)
                                          n 0

                     Stefan Mangold - ComNets Aachen University                                     39
          Condition for Cooperation
It is more efficient to cooperate instead of defect (instead
of playing best response), if…
                                    n  n' 1                     
        
          i k      i
                VCC         i
                             VDC               
                                                  i k      i
                                                        VCD                
                                                                               i k      i
                                                                                     VCC
  k n                                  k n 1               k n  n' 1

It depends on the discounting factor (importance/shadow
of future) if mutual support is achievable:
                                           i     i
                                          VCC  VDC
                                i         i    i
                                          VCD  VDC

The more important the future is, the more likely is the
establishment of cooperation
For example, CCHCs will interact for many superframes


                       Stefan Mangold - ComNets Aachen University                           40
              Dependence on Discounting Factor
              1.5                                                                              1.5
                            i = 1                                                                                                                 i = 0.8
                                              Future counts
of player i




                                                                                 of player i
                                                            ViCOOP( i=1)
               1                                                                                1
                                                                                                     ViCOOP( i=0.8)
        MSG




                                                                                         MSG
Vi




                                                                                 Vi
              0.5                                                                              0.5
                       1     2    3     4   5      6    7    8         9    10                           1     2    3     4   5      6    7    8       9      10
                           stages of punishment through player -i                                            stages of punishment through player -i


              1.5                                                                              1.5
                                                                                                                                                   i = 0.6

                                                                                                                 Future is less important
of player i




               1                                                                 of player i    1
                    ViCOOP( i=0.75)                                                                 ViCOOP( i=0.6)
        MSG




                                                                                         MSG
Vi




                                                                                 Vi




                                                           i = 0.75
              0.5                                                                              0.5
                       1     2    3     4   5      6    7    8         9    10                           1     2    3     4   5      6    7    8       9      10
                           stages of punishment through player -i                                            stages of punishment through player -i

                                                  Stefan Mangold - ComNets Aachen University                                                                       41
                         Wrap Up
 There is always a Nash equilibrium in the single stage
 game
 If the outcome of the Nash equilibrium is not satisfying, a
 player may attempt to punish the opponent, for
 establishment of mutual support
 Depending on the behaviors of the CCHCs (the interacting
 players), and their requirements, cooperation can be
 achieved
 QoS can be supported if cooperation is established

         QoS Support in legacy 802.11?  no!
 QoS Support in 802.11e EDCF?  yes, but no guarantee!
QoS Support with 802.11e HCF?  not in unlicensed bands!
 QoS Support with shared radio resources?  with mutual
                       support: yes!
                Stefan Mangold - ComNets Aachen University     42
                     Conclusions
IEEE 802.11e EDCF will provide basic means for QoS
support
The controlled medium access of HCF (polling) cannot
support QoS in unlicensed frequency bands
New analytical model for EDCF is developed
    allows to predict and control QoS
New approach for coexisting radio networks
    may help radio networks operating in unlicensed bands to
    support QoS
Results will be used in …
    Contributions to IEEE 802.11e
    IEEE 802.19 coexistence discussions
    Spectrum etiquette development at Wi-Fi alliance
    Development of Spectrum Agile Radios (DARPA)

                Stefan Mangold - ComNets Aachen University     43
Backup Slides
                       Architecture
Infrastructure Basic Service Set (BSS)
    one station is the access point
Independent Basic Service Set (IBSS)
    ad-hoc                                            DS
                802.x LAN
                 via Portal

                            Wired Station                                Wired Station
                          (Access Point, AP)                          (Access Point, AP)

                                                BSS                                         BSS




                   Wireless                Wireless              Wireless              Wireless
                   Station                 Station               Station               Station
                              Wireless                                      Wireless
                              Station                                       Station




                                Wireless
                                                      Wireless
                                Station
                                                      Station

                                           Wireless
                                 IBSS      Station

                Stefan Mangold - ComNets Aachen University                                        45
                 Medium Access - Example
Station 1 initiates frame exchange first
Other stations set the Network Allocation Vector (NAV)
Distributed approach  difficult for station to support QoS
                                                                                                new random
                                           S                                           S
                                                                                                  backoff
                                 random I                                              I
                                                                                                 (10 slots)
station 1                         backoff F CTS                                        F ACK
                                 (7 slots) S                                           S
                        D                         S                                             D    station
station 2               I            RTS          I                DATA                         I    defers
                        F                         F                                             F
                        S                         S                                             S
                                random                                                               remaining
                NAV
                                back-off                                                               backoff
station 3       reset
                                (9 slots)                                                             (2 slots)
                                                                                                                      ACK

                        D                                                                       D                 S
station 4               I                                                                       I       DATA      I
                        F                                   stations set NAV upon               F                 F
              S         S                                                                       S                 S
                                                                 receiving RTS
              I
station 5     F ACK
              S

station 6                      station 3
                DATA          defers, but
                            keeps backoff
                                            station 6 sets NAV upon receiving CTS,     NAV                        NAVs
                             counter (=2)
                                                this station is hidden to station 1   updates
       NAV (timer)
       transmission                                                                                 time

                                Stefan Mangold - ComNets Aachen University                                                  46
Multiple Backoff Entities per Station
       legac y 802.11 station                IEEE 802.11e station with four backoff entities:
      with one backoff entity:
                                                       8 priorities 0 - 7 according to 802.1D are
                                                        mapped to 4 Access Categories (ACs)
                                                  7        6        5        4       3        2         1          0




          one priority                             4 Access Categories AC0 - AC3 representing 4
                                                    priorities, with 4 independent backoff entities
                         backoff               higher priority                                    lower priority
                          entity                                                                                   backoff
                                                 AC3               AC2             AC1                 AC0
                                                                                                                    entity




            backoff:                          backoff:           backoff:         backoff:         backoff:
             DIFS                  PF[AC] not AIFSN[3]           AIFSN[2]         AIFSN[1]         AIFSN[0]
               15                    part of  CWmin[3]           CWmin[2]        CWmin[1]          CWmin[0]
             1023                   802.11e   CWmax[3]           CWmax[2]        CWmax[1]          CWmax[0]


                                              upon parallel access at the same slot, the higher priority AC
                                               backoff entity transmits, the other backoff entity/entities act
                                                                 as if a collision occured
                                   AIFSN = 1,2,3…                                transmission
                 transmission      AIFS = SIFS + aSlotTime x AIFSN




                         Stefan Mangold - ComNets Aachen University                                                          47
                           Markov Chain
State transition probabilities

  P s  t  1  AC |s  t   slot   Pslot , AC ,   AC  H , M , L,
  P s  t  1  C |s  t   slot   Pslot ,C ,
  P s  t  1  slot  1|s  t   slot   Pslot ,slot 1 , slot  1 CWmax  1


Stationary distributions
              lim P s  t   AC   p AC ,         AC  H , M , L ,
             t 
              lim P s  t   C   pC ,
             t 
              lim P s  t   slot   pslot , slot  1 CWmax  1
             t 




                       Stefan Mangold - ComNets Aachen University                   48
                    Allocation Process (Example)
            Two single stage games (two superframes):
                                                                                                beacons
                                                                                               (at TBTTs)
                                collision                collision         collision
                                                                                                player 1
TXOPs




                                                                                                player 2


                                                                                                player 3
                                                                                                 (EDCF)
        0      40    80   120          160      200       240        280    320        360   400
                                             time [ms]




            Two players interact with each other
            A third player models the EDCF background traffic
            For analysis, a formal description of this process is needed




                           Stefan Mangold - ComNets Aachen University                                   49
                     Strategy: Best Response

pl1     1         required
                  observed       as both players play their                pl2 1                   now demanding high thrp. when converging into NE

                  demanded       best responses, the demands                                                     this player gains from playing
                                 converge into Nash equilibrium (NE)                                             the best response
       0.6                                                                        0.6
                                                                                                                                                       
  1




                                                                            2
                                           apprx.                                                                                    required
                          sim.                                                                                 sim.
                    observed thrp. decreases, because now the                             apprx.                                     observed
                    opponent player 2 plays its best response                                                                        demanded
        0                                                                          0

       0.1                                       required                         0.1                                             required
                                                 observed max                                                                     observed max
                                                 demanded                                          apprx.
                                                                                                                                  demanded
                                                                                         sim.
                                                                                                                                                       
 1




                                                                           2
      0.04
                              sim.                    demanding NE              0.023
                                             apprx.                                                                                  demanding NE
        0                                                                          0
        0.2   1    1.8     2.6 3.4 4.2 5 5.8           6.6      7.4                0.2     1       1.8     2.6 3.4 4.2 5 5.8             6.6      7.4
                         time (s), SFDUR = 200ms                                                         time (s), SFDUR = 200ms


 Best Response: adapt demand to achieve highest outcome
 (myopic competition)
 Action profile (demand) converges to NE

                                       Stefan Mangold - ComNets Aachen University                                                                           50
                           Strategy: Cooperation

pl1     1
                        apprx.
                        (not visible)
                                                  required
                                                  observed
                                                                    pl2     1
                                                                                                                        required
                                                                                                                        observed
                                   sim.
                                                  demanded                                                              demanded
       0.6                                                                 0.6
                                                                                                                                       
  1




                                                                     2
                                                                                  sim.        apprx.

        0                                                                   0

       0.1                                      required                   0.1                                       required
                                                observed max                                                         observed max
                        sim.                    demanded                                 apprx.                      demanded
                               apprx.                                                              sim.
                                                                                                                                       
 1




                                                                    2
      0.04
                                                                         0.023

        0                                                                   0
        0.2   1   1.8     2.6 3.4 4.2 5 5.8         6.6   7.4               0.2     1       1.8     2.6 3.4 4.2 5 5.8       6.6   7.4
                        time (s), SFDUR = 200ms                                                   time (s), SFDUR = 200ms


 Cooperation: reduced demand, shorter resource
 allocations
 Now both players achieve higher outcomes (next page…)
                                          Stefan Mangold - ComNets Aachen University                                                        51

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:5
posted:10/10/2011
language:English
pages:52