Академический Документы
Профессиональный Документы
Культура Документы
h-r,.ivoril~-PL.r~~rNtrrrrce
Asn$.sk;
802.1I
MAC;
Qurrrc
~ ~ l l I l l ~ , i C . T'
I.
1NTIIC)DUCIION
II.
A.
BACKGROUND
2586
The 14" IEEE 2003 InternationalSymposium on Persona1,lndoorand Mobile Radio Communication Proceedings
E.
c r u i i s r i i i s s i ~procediire
~;~
MAC L A Y E R
Service T i m
PI.
B.
p, = l - [ l - ( l - p o ) r ) l ' - '
(1)
where po is the probability that there are no packets ready to
transmit at the MAC layer in the wireless station under
consideration, and T is the packet transmission probability that
the station transmits in a randomly chosen slot time given that
the station has packets to transmit at the M A C layer. Let W be
the iiiinimum value of contention window size. Since the
contention window is reset after the maximum a times o f
retransmissions as defined in the protocols [ I O ] > following
similar procedure used in [6] and [9], we can obtain
.a>m
I-~~:"~p~~~~~,'(2p')'+cv(I-2"'p'"")
.(a
(3)
= (,1 - [)((I
- [i"""'-''
+p.-I)
Then pc - P,,, is the probabilip that there are collisions
among other (11-1) stations (or neigh!,ors).
c,,~.
2(1- p,"" )
2587
The i4IhIEEE 2003 International Symposium on Personal,lndoor and Mobile Radio Communication Proceedings
;:
-\
/f,( (Z)
/ ( 2 ' / V 1,
J
l./ly,, 1Z )'
/.I,
z) =
n;,=,,
/./l,((z).(O
/I(%)= ( I
-/I,
(05 i 5 I N )
Figire I .
( m< i 5 a)
02
;s
.(5)
li
0 15
)Z'.,"
,=I#
B(z)=c,:"P(7:= i ) Z ' ,
(6)
d"iiioiiient of Ti by differentiation
(hence the inean value and the variance). Herc the unit o f
slot.
r, is
"8"s
r,,,
E. Simularron Resulrs
In our simulation study, we use ns-2. The wireless channel
capacity is set to ZMbps. Packet length i s 1000 bytes. and the
maximum queue length i s 50. We use different number of
mobile stations in a rectangular grid with dimension 2OOm x
200m to simulate the Wireless LAN.
The mobile stations are randomly moving arid have the
same rate o f packet input. Our simulation indicates the
distribution o f M A C layer service time i s independent o f the
packet input distribution whether it is deterministic or Poisson
distributed. It mainly depends on the total traffic in the
network before saturation and on the number of mobile stations
after saturation.
Fig.3 shows the simulation results o f the M A C layer service
time in the network with 17 mobile stations and total traftic of
0.8 and 1.6 Mbps respectively.
It displays the same
characteristics with that obtained from the theoretical result.
2588
The 14" IEEE 2003 International Symposium on Persona1,lndoor and Mobile Radio Communication Proceedings
IV.
.A
MOIXI.INC:ANIIANALYSIS
Q~JI:UI:IN(;
P~ohletn,li,rinuliilioii.
17. So we
can compute p, under different values ofj.'and n with the help
of (I)and (2) using some recursive algorithm. Thus we can
obtain the distribution of M A C service time at different offered
load verified by simulation results in Fig. 3.
D.
(7)
P,, = Pr{Xn*.,=j 1 X,,= i ) ,
where !,'A is the number of packets seen upon the nth departure.
Per/ormonce Melrics
E, Throughput
Ifwe know the blocking probability pH,then the throughput
Scan be computed easily by
s = ( I 7 PS 10 - P'""
To obtainp,,. we define
'
where p/
i s the
transmission failure.
,~,.
>
(14)
F. Results
where i'
is the average arrival rate. We can easily obtain
10
...
io like
2589
''
The 14'O IEEE 2003 International Symposium on Persona1,lndoor and Mobile Radio Communication Proceedings
the MAC layer service time. so they show the similar change in
the liyui-es. Average queue lengh is almost zero at the nonsaturated status and reaches alniost maximum length at the
saturated status. Average waiting tiine for each packet in the
queue alniost equals zero at non-saturated status and reaches
se\wal seconds at the saturated status. Queue blocking
prohahilit). is zero at the lion-saturated status when traffic load
is low. and linearly increases with the offered load at the
saturated status. Throughput linearly increases with the offered
load at the non-sdturated status and inaintains a constant value
with different total number of transmitting stations at the
saturated S I ~ I I U S . The packet discarding probability at MAC
l a y is iiiucli smaller than the queue blocking probability.
$ 03
".ti
".9
n=5
"~
.,.~....I
n=33
g o d
'
~
REFERENCES
02
V.
CoNci.usiow
[7j
[XI
47-54
191 14. Wu. Y. Peng. K. Long. et ill. 'Performance of rcliabie
transpotl protocol ovc
E 802.1 I wireless LAN: analysis and
cnhancrmcnt." Proc. I
INFOCOM'2002. June. 2002.
l l 0 l IEEE standard for Wireless L A N Medium Access C m r o i ( M A C ) and
l'llyriciil Layer (PHY) specifications. ISOIlEC 8802-1 I: 1999(E), Aug.
1999
11 I1 K .
2002
2590
3rd cd..