A COMPARATIVE STUDY OF IEEE 802.11 MAC ACCESS
MECHANISMS FOR DIFFERENT TRAFFIC TYPES
Mohammad Saraireh, Reza Saatchi, Samir Al-khayatt, Rebecca Strachan
Sheffield Hallam University, Pond Street,Sheffield, UK
Keywords: Quality of Service (QoS), IEEE 802.11 Medium Access Control (MAC) Protocol, Network Performance.
Abstract: The fast growth and development of wireless computer networks and multimedia applications make the
Quality of Service (QoS) provided to their transmission an important issue. This paper aims to investigate
the impact of varying the number of active stations on the network performance. This was carried out using
different data rates. The investigations also considered both MAC protocol access mechanisms, i.e. the basic
access and the Request To Send / Clear To Send (RTS/CTS). The effect of traffic type i.e. Constant Bit Rate
(CBR) and Variable Bit Rate (VBR) traffics was also examined. The findings revealed that in large
networks (larger than 15 stations), the RTS/CTS access mechanism outperformed the basic access
mechanism since the performance of the latter was more sensitive to the increase and decrease of the
number of active stations. Increasing the data rate improved the network performance in term of delay and
jitter but it degraded the network performance in term of channel utilisation and packet loss ratio.
1 INTRODUCTION
Wireless systems are increasingly used for
transmitting different type of applications such as
voice, video and data. Wireless transmission
requires a controller to manage accessing the
medium in a fair and suitable manner and to share
the resources. Random transmission may lead to
incomprehensible or unpredictable results.
Therefore, a controller for accessing and sharing the
resources is an essential tool for achieving a
successful transmission process between the
communication parties.
The Medium Access Control (MAC) protocol in
wireless networks controls access to the shared
medium by applying rules and procedures that
permit the communication pairs to communicate
with each other in an efficient and fair manner.
The IEEE 802.11 standard defines two
coordination functions (IEEE, 1999). They are
Distributed Coordination Function (DCF) and Point
Coordination Function (PCF). The focus of this
study is the DCF that is part of the Carrier Sense
Multiple Access with Collision Avoidance
(CSMA/CA).
Under DCF protocol, data packets are transferred
using two mechanisms. The main mechanism is a
two-way handshaking process which is called basic
access mechanism. The optional or alternative
mechanism is called RTS/CTS access mechanism
that based on the exchange of RTS and CTS
messages before data packets transmission.
RTS/CTS access mechanism is used to reserve
the channel before data transmission. Under DCF,
all stations in the same Basic Service Set (BSS) have
to compete between each other to gain access to the
medium. The competition between stations is
controlled by different parameters of the physical
layer (PHY) and the MAC sub-layer. The
parameters include the Inter Frame Space (IFS) i.e.
time period between the transmission of frames,
Contention Window (CW), and backoff mechanism
that randomises instants at which stations are
attempting to access the channel.
All these parameters play important roles on the
network performance through their effect on the
degree of competition between the active stations
within the same BSS. Consequently, an increase in
the number of active stations in the BSS increases
the degree of competition which in turn increases the
probability of collisions. As a result of that, an
increase in the number of stations has an obvious
impact on the network performance.
28
Saraireh M., Saatchi R., Al-khayatt S. and Strachan R. (2005).
A COMPARATIVE STUDY OF IEEE 802.11 MAC ACCESS MECHANISMS FOR DIFFERENT TRAFFIC TYPES.
In Proceedings of the Second International Conference on e-Business and Telecommunication Networks, pages 28-35
DOI: 10.5220/0001420400280035
Copyright
c
SciTePress
2 RELATED WORK
The variation of the number of active stations in
IEEE 802.11 DCF protocol has been investigated in
several studies by both simulation tools and
mathematical models. An analytical model was
proposed to analyse DCF operation and compute the
saturated throughput performance through
employing Markov chain models (Bianchi, 2000).
This proposed model considered a finite number of
stations with ideal channel conditions. The results
obtained in this paper showed that the performance
of the basic access mechanism depends on the MAC
parameters mainly contention window minimum and
number of wireless stations in the wireless networks.
On the other hand, the results showed that the
RTS/CTS access mechanism is marginally
dependent on the system parameters. In another
study the capacity of the medium was investigated
by developing a mathematical model that calculates
the DCF throughput and the packet virtual
transmission time (Cali, 2000).
IEEE 802.11 CSMA/CA protocol over wireless
channel was investigated in (Kleinrock, 1975). They
provide an analysis for the channel performance
during the up-time of unstable channel. They
showed that CSMA theoretically exhibits behaviour
similar to ALOHA. In (Haitao, 2002) a scheme
named DCF+, which is compatible with DCF; to
enhance the performance of reliable transport
protocol over WLAN was proposed. Moreover, the
impact of increasing the number of stations on the
saturated throughput and delay in DCF and in the
proposed scheme DCF+ was investigated. Their
results revealed that increasing number of stations
has an obvious impact on the network performance.
In (Sweet, 1999), throughput performance
measures for varying number of stations in
CSMA/CA were presented. They showed that the
RTS/CTS access mechanism achieved higher
throughput for CBR traffic when the number of
stations increased above 10 stations. Their results
also showed that higher transmission speeds yielded
lower average throughput results.
Changing the number of active stations has an
obvious impact on achieving QoS over wireless ad-
hoc networks. This is due to the increase of collision
probability over the medium. Also varying the data
rate has a considerable impact on the average end-
to-end delay and jitter. These parameters have
critical impact on the transmission of multimedia
applications.
An aim of this study is to investigate the impact
of increasing the number of active stations and data
rate on the network parameters. In particular, on the
QoS parameters, throughput, end-to-end delay, jitter,
and data packets drop. The performance of MAC
protocol access mechanisms for CBR and VBR
traffics was analysed.
This paper is organised into five sections. In the
next section, the basics of the IEEE 802.11 MAC
protocol are introduced. The experimental procedure
is introduced in section 4. The findings and
discussions are presented in section 5. The
conclusion and future work is presented in section 6.
3 IEEE 802.11 MAC PROTOCOL
The IEEE 802.11 standard (IEEE, 1997) specifies a
CSMA/CA protocol. In CSMA/CA, when a station
has a packet to send, it first listens to the medium to
ensure no other transmission is currently taking
place. If the channel is idle, it then transmits the
packet. Otherwise, it picks a random "backoff
interval" which determines the period of time the
station has to wait until it is allowed to transmit its
packet. The selection of the random number of the
backoff time is based on a binary exponential
backoff algorithm. The competing stations select a
random number between 0 and CW-1 with equal
probability. If the data packet is successfully
transmitted, the backoff counter of the transmitted
station will reset and then the station starts to
compete with the other stations for accessing the
medium. During the idle period of the channel, the
transmitting station decrements its backoff counter.
When the backoff counter reaches zero, the station
transmits the packet as shown in Figure 1. During
the busy period the station suspends its backoff
counter. After successful receiving a packet, the
receiving station replies with a positive
acknowledgement (ACK) after waiting for a Short
Inter Frame Space (SIFS) period. If an ACK is not
detected within a SIFS period after the packet
transmission, the transmission is assumed to be
unsuccessful, and a retransmission is scheduled
according to the specified backoff rules. The
unsuccessful transmission is due to collision over the
link. If a collision occurs CW will be doubled until
reaching the maximum value CW
max
= 2
m
(CW
min
+
1) – 1, where m is the number of retransmission
attempts.
The RTS/CTS access mechanism is mainly used
to minimize the amount of time spent when a
collision occurs since collision occurs in these short
messages.
A COMPARATIVE STUDY OF IEEE 802.11 MAC ACCESS MECHANISMS FOR DIFFERENT TRAFFIC TYPES
29
Before commencing the transmission of a data
packet, the source station sends a short control
frame, called RTS, declaring the duration of the
forthcoming transmission. When the destination
station receives the RTS frame, it replies with a CTS
frame after SIFS interval, with the duration of the
future transmission. Upon hearing RTS and CTS, all
other stations in the vicinity of the sender and the
receiver update their Network Allocation Vectors
(NAV). This process reserves the medium for the
sending station. Thus, all stations in the
neighbourhood of the sender and receiver defer their
transmissions and receptions to avoid collisions.
After the successful RTS/CTS exchange, the source
station transmits the data packet then the receiver
responds with an ACK frame. Figure 2 depicts the
time line of the RTS/CTS access mechanism.
4 EXPERIMENTAL PROCEDURE
To analyse the impact of varying the number of
active stations and data rate on the network
performance for both the MAC protocol access
mechanisms and the two different traffic types a
number of simulation studies were carried out using
the network simulator package (ns2) version 2.27
(ns2). The studies were carried out under different
scenarios and they were based on the QoS
parameters; throughput, delay, delay variation, and
packet loss.
The performance of the IEEE 802.11 MAC
protocol was investigated when the number of active
stations in the same BSS was increased. Two
different channel data rates were chosen for data
packet transmission; low data rate equal to 2Mbps
and high data rate equal to 11Mbps. While the
control frames were transmitted at data rate equal to
1Mbps. IEEE 802.11b standard was used since it
offers multi data rates. The protocol parameter
settings were as shown in Table 1. A random
topology with 20 stations was adopted when all
stations were located in the same BSS. The network
was offered by 100% of offered load every time the
simulation run. Each connection was specified as a
source - destination pair in which the number of
connections was varied each time the network
simulation was run. The simulation was carried out
for CBR and VBR traffic at both MAC protocol
access mechanisms. The CBR traffic had fixed
packet size (1280-byte) while the VBR traffic had
variable packet size and variable interval (mean
packet size 3993 bytes and 2541 bytes standard
deviation).
All nodes were arranged in a random topology
with area of 200x200 metre with the help of random
way point model, and the same model was used for
all the simulations. Throughout the simulations, all
nodes were within range of each other and there
were no hidden terminals occurrences. Each scenario
was run for 10 times. The results were the mean
value for simulations. Each simulation was run for
duration of 100 seconds using Ad-hoc On-demand
Distance Vector (AODV) as the routing protocol
since it has proven to be efficient as opposed to
proactive protocol in Mobile Ad-hoc Networks
(MANET) (Broch, 1998).
Figure 1: Timeline of Basic access mechanism in DCF.
Figure 2: Timeline of RTS/CTS access mechanism in DCF.
ICETE 2005 - WIRELESS COMMUNICATION SYSTEMS AND NETWORKS
30
5 RESULTS AND DISCUSSIONS
This section outlines performance evaluation of the
Distributed Coordination Function (DCF) that is a
part of the IEEE 802.11 standard. It demonstrates
through simulations the performance of the IEEE
802.11 MAC protocol when the number of stations
is varied. Further, the impact of this variation on the
QoS parameters is analysed. A comparison of the
access methods provided by the IEEE 802.11 MAC
protocol is carried out and comments are made as to
when each should be employed.
5.1 Average Throughput
An increase in the number of contending stations in
the same BSS causes more collisions and as a result
more channel bandwidth is wasted. This wastage of
bandwidth causes a reduction in the achieved
throughput for both MAC protocol access
mechanisms
.
Figures 3a and 3b show the relationship between
the active stations and the channel utilisation
(channel utilisation is the ratio of the received bits to
the channel data rate). When the number of active
stations was increased, the channel utilisation
slightly declined when the RTS/CTS access
mechanism was used compared to the basic access
mechanism. In RTS/CTS mechanism, collisions only
involve control frames which are relatively small in
size compared to data packet sizes, hence the
bandwidth wasted in collisions is less than the basic
access mechanism. This explains the slight rate of
decrease in the channel utilisation curve when the
RTS/CTS access mechanism was used.
If there are few stations in the network, (i.e. less
than 10 stations), the RTS/CTS access mechanism
provided a lower channel utilisation and lower
average throughput. This was due to the overhead
introduced by the control frames RTS and CTS. The
impact of this overhead on the average throughput
became very small when data packet sizes was very
large (above 2000 bytes) as shown Figure 4. At
small packet sizes, the basic access mechanism
outperformed the RTS/CTS access mechanism due
to the impact of the overhead, while at large packet
sizes, the RTS/CTS access mechanism outperformed
the basic access mechanism since the size of RTS
and CTS is very small compared to data packet
sizes.
With regard to channel data rate, low data rate
(2Mbps) achieved better channel utilisation than
high data rate (11 Mbps). This is because in low data
rate the data packets were sent at 2 Mbps while the
headers and control frames were sent at basic rate (1
Mbps). The two data rates (low data rate and basic
rate) are relatively close to each other which resulted
in better channel utilisation. At high data rate, the
data packets were sent at 11 Mbps while the headers
and control frames were sent at 1 Mbps, the
difference here was relatively high compared to low
data rate which resulted in a high rate of reduction in
the channel utilisation. In this case, the transmission
of headers and control frames caused a bottleneck
when data packets were sent at high data rate.
The channel utilisation was degraded for CBR
and VBR traffics. For CBR traffic, the reduction
over an increase from 1 to 20 stations was 4.3% at
11 Mbps and 10.5% at 2 Mbps when the basic
access mechanism was used. When the RTS/CTS
access mechanism was used the reduction was
slightly smaller, it was 2.4% at 11 Mbps and 3.5% at
2 Mbps. For VBR traffic, the channel utilisation
degraded by 4.8% at 11 Mbps data rate and by
11.5% at 2 Mbps data rate when the basic access
mechanism was used. When the RTS/CTS access
mechanism was used the reduction in the channel
utilisation was 1.8% at 11 Mbps and 3.5% at 2
Mbps. The results obtained indicate that the channel
utilisation was degraded for CBR and VBR traffic in
both; the RTS/CTS and the basic access
mechanisms, but the RTS/CTS access mechanism
provided a smaller rate of decrease in the channel
utilisation when the number of stations was
increased. Also high data rate (11 Mbps) provided a
slight decrease in the channel utilisation compared to
the low data rate (2 Mbps).
Parameter Value
Data Rate 2, 5.5, 11 Mbps
Basic Rate for broadcast 1 Mbps
DIFS 50 µsecs
SIFS 10 µsecs
CWmin 31
CW max 1023
Slot time 20 µsec
Short Retry Limit 7
Long Retry Limit 4
Table 1: IEEE 802.11b Parameter (ORiNOCO) settings.
A COMPARATIVE STUDY OF IEEE 802.11 MAC ACCESS MECHANISMS FOR DIFFERENT TRAFFIC TYPES
31
30
35
40
45
50
55
60
65
70
75
80
85
2 4 6 8 10 12 14 16 18 20
Number of Stations
Channel utilisation [percent]
Channel utilisation at [11Mbps data rate] / basic
Channel utilisation at [2 Mbps data rate] /basic
Channel utilisation at [11Mbps data rate] /RTS/CTS
Channel utilisation at [2 Mb data rate] /RTS/CTS
30
35
40
45
50
55
60
65
70
75
80
85
2 4 6 8 10 12 14 16 18 20
Number of stations
Channel utilisation [percent]
Channel utilisation at [11 Mb-data rate] /basic
Channel utilisation at [11 Mb-data rate] / RTS/CTS
Channel utilisation at [2 Mb-data rate] /basic
Channel utilisation at [2 Mb-data rate] /RTS/CTS
0.40
0.55
0.70
0.85
1.00
1.15
1.30
1.45
1.60
1.75
0 500 1000 1500 2000 2500 3000 3500 4000
Packet size [bytes]
Avearge throughput [Mbps
]
Average throughput [Mbps] /basic
Average throughput [Mbps] /RTS
5.2 Average Delay
The packet delay from and end-to-end should not
exceed 400 ms for time sensitive applications in
order to achieve the required QoS (Coverdate,
2000). As shown in Figures 5a, 5b, 5c and 5d, low
data rate (2Mbps) in both MAC access mechanisms
does not meet this QoS requirement if the number of
active station was increased to more than 4 stations.
A high data rate (11 Mbps) achieved better
performance (small values of average delay). The
average delay was slightly increased which met the
QoS requirements up to 10 stations and then started
to exceed the limit as the number of active stations
was increased
.
Because of the strict delay and jitter
requirements for multimedia applications (CBR and
VBR traffics), the time interval between the
packet transmissions has to be within a given period.
This can be obtained by assigning small values of
CW
min
and CW
max
for these applications. In this
study The IEEE 802.11 and IEEE 802.11b protocols
were used, and their CW
min
and CW
max
were kept
at the default values (31 and 1023 respectively) for
the
pairs of communication, therefore, their delay
and jitter values were increased at low data rates and
slightly increased at high data rates
.
At high data rate (11 Mbps) with the basic access
mechanism, the average delay was reduced by 69%
and 66% compared with low data rate for CBR and
VBR traffics, respectively. When the RTS/CTS
access mechanism was used, the average delay at
high data rate was also reduced by 58% and 63% for
CBR and VBR traffics respectively.
The values of average delay in both MAC access
mechanisms were located outside the desired range
of QoS (150 ms for high QoS and 400 ms the
minimum limit) when low data rate (2 Mbps) was
used (Coverdate, 2000). Conversely, high data rate
(11 Mbps) can provide acceptable QoS requirements
in term of average delay.
5.3 Average Jitter
One of the major roles of QoS is to keep delay, jitter
and packet loss for the transmitted applications
within the acceptable range (Coverdate, 2000). For
instance, to achieve high QoS for multimedia
applications, the average jitter should not exceed 20
ms.
The average jitter increased as the number of
active stations was increased. In other words, as the
number of stations was increased; the probability of
collisions increased due to a high degree of
competition between stations. This in turn forced the
MAC protocol to retransmit the collided packets. If
the collided packets were successfully received at
the destinations, they experienced delay variation,
and this variation depended on the number of packet
retransmissions.
Figure 4: Average throughput at Basic and RTS/CTS
access mechanisms when the packet size increased
.
(a) (b)
Figure 3: Channel utilisation for CBR and VBR traffic at two different data rates and at two MAC protocol
access mechanisms. (a) Channel utilisation for CBR. (b) Channel utilisation for VBR.
ICETE 2005 - WIRELESS COMMUNICATION SYSTEMS AND NETWORKS
32
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00
2.20
2 4 6 8 10 12 14 16 18 20
Number of stations
Average end-to-end delay [second]
Average End-to-End Delay
at [11 Mbps-data rate]
Average End-to-End Delay
at [2 Mb-data rate]
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00
2.20
2.40
2 4 6 8 101214161820
Number of stations
Average end-to-end delay [second]
Average End-to-End Delay
at [11 Mbps-data rate]
Average End-to-End Delay
at [2 Mb-data rate]
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2 4 6 8 10 12 14 16 18 20
Number of stations
Average end-to-end delay [second]
Average End-to-End Delay
at [11 Mbps-data rate]
Average End-to-End Delay
at
[
2 Mb-data rate
]
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2468101214161820
Number of stations
Average end-to-end delay [second]
Average End-to-End
Delay [11Mb-data rate ]
Average End-to-End
Delay [2 Mb-data rate]
0.00
0.01
0.02
0.03
0.04
0.05
0.06
2 4 6 8 10 12 14 16 18 20
Number of stations
Average jitter [second]
Average Jitter at
[11 Mbps-data rate]
Average Jitter at [2
Mb-data rate]
0.00
0.01
0.02
0.03
0.04
0.05
0.06
2 4 6 8 10 12 14 16 18 20
Number of stations
Average jitter [second]
Average Jitter at [11
Mbps-data rate]
Average Jitter at [2
Mb-data rate]
0.00
0.01
0.02
0.03
0.04
0.05
0.06
2 4 6 8 10 12 14 16 18 20
Number of stations
Average jitter [second]
Average Jitter at [11
Mbps-data rate]
Average Jitter at [2
Mb-data rate]
0.00
0.01
0.02
0.03
0.04
0.05
0.06
2 4 6 8 10 12 14 16 18 20
Number of tations
Average jitter [second]
Average Jitter
[11 Mb-data rate]
Average Jitter [2
Mb-data rate]
As shown in Figures 6a, 6b, 6c and 6d, the
transmission of data packets with high data rate
(11Mbps) had a noticeable positive impact on the
achieved value of average jitter. High data rate
resulted in small values of average jitter. This was
because the transmission time of data packets at high
data rate was smaller.
The results obtained at 11 Mbps indicated that
the values of average jitter for CBR and VBR traffic
in both MAC access mechanisms were kept within
the acceptable range of QoS (less than 20 ms),
where as low data rate resulted in large values of
average jitter (more than 20 ms).
5.4 Packet Loss
In this study the packet loss was due to collisions,
especially when the MAC retry limit exceeded and
buffer overflow. Figures 7 and 8 show that
the performance was downgraded with the increase
in the number of stations in the same BSS.
It is well-known that high transmission rates
have a lower Signal to Noise Ratio (SNR) than low
data rates (Thruong, 2003). Therefore, at high data
rates, the probability that a packet can not be
received correctly by the destination is high.
0
5
10
15
20
25
30
2 4 6 8 10 12 14 16 18 20
Number of stations
Data packets drop due
to collisions [percent]
Percentage of data packet drop due t o
collision at [11 Mbps-data rate]
Percentage of data packet drop due t o
collision at [2 Mb-data rate]
0
5
10
15
20
25
2 4 6 8 10 12 14 16 18 20
Number of stations
Data packets drop due to collisions
[percent]
Percentage of dat a packet drop due to
collision at [11 Mbps-data rate]
Percentage of dat a packet drop due to
collision at [2 Mb-data rate]
0
10
20
30
40
50
60
2 4 6 8 10 12 14 16 18 20
Number of stations
Control frames drop due to collision
[percent]
Percentage of data packet drop due
to collision at [11 Mbps-data rate]
Percentage of data packet drop due
to collision at [2 Mb-data rate]
0
5
10
15
20
25
30
35
2 4 6 8 10 12 14 16 18 20
N
umber of stations
Control frames drop due to
collisions [percent]
Percentage of data packet drop due
to collision [11 Mb-data rate]
Percentage of data packets drop
due to collision [2 Mb-data rate]
(a): Average end-to-end
delay for CBR / basic.
(b): Average end-to-end
delay for VBR / basic.
(c): Average end-to-end
delay for CBR / RTS.
(d): Average end-to-end
delay for VBR / RTS.
Figure 5: Average end-to-end delay vs. number of stations.
(a): Average jitter for
CBR / basic.
(b): Average jitter for
VBR / basic.
(c): Average jitter for
CBR / RTS.
(d): Average jitter for
VBR / RTS.
Figure 6: Average jitter vs. number of stations.
(a): Data packet drop due to
collision for CBR/basic.
(b): Data packet drop due
to collision for VBR/basic.
(c): Control frame drop due
to collision for CBR/RTS.
(d): Control frame drop due
to collision for VBR/RTS.
Figure 7: Percentage of Collision drop vs. Number of stations.
A COMPARATIVE STUDY OF IEEE 802.11 MAC ACCESS MECHANISMS FOR DIFFERENT TRAFFIC TYPES
33
0
10
20
30
40
50
60
2 4 6 8 10 12 14 16 18 20
Number of stations
Percentage of data packets drop due
to buffer overflow [percent]
Percentage of data packet drop due to
buffer overflow at [11 Mbps-data rate]
Percentage of data packet drop due to
buffer overflow at [2 Mb-data rate]
0
10
20
30
40
50
60
2 4 6 8 10 12 14 16 18 20
Number of stations
Percentage of data packets drop due
to buffer overflow [percent]
Percentage of dat a packet drop due to
buffer overflow at [11 Mbps-data rate]
Percentage of dat a packet drop due to
buffer overflow at [2 Mb-data rate]
0
10
20
30
40
50
60
70
2 4 6 8 10 12 14 16 18 20
Number of stations
Percentage of data packets drop due to
buffer overflow [percent]
Percentage of data packet drop due to
buffer overflow at [11 Mbps-data rate]
Percentage of data packet drop due to
buffer overflow at [2 Mb-data rate]
0
10
20
30
40
50
60
70
2 4 6 8 10 12 14 16 18 20
Number of stations
Percentage of data packets drop due to
buffer overflow [percent]
Percentage of data packet drop due to
buffer overflow [11 Mb-data rate]
Percentage of dat a packet s drop due
to buffer overflows [2 Mb-data rate]
The results obtained showed that the drop due to
buffer overflow is relevant in the total loss ratio only
with low data rate if the number of active stations
was increased, while with high data rate the main
cause of packet loss was the collisions (MAC retry
limit) as shown in Figures 7 and 8
As shown in Figures 8a, 8b, 8c, and 8d, at high
data rates the packet loss ratio was larger by 30%
and 23% than the obtained values at low data rates
for CBR and VBR traffics, respectively when the
basic access mechanism was used. When the
RTS/CTS access mechanism was used, the packet
loss ratio was larger by 36% and 35% than the
obtained values at low data rates for CBR and VBR
traffic, respectively.
In this scenario, the basic mechanism
outperformed the RTS/CTS mechanism in term of
packet loss when the number of stations was small.
For large networks, the RTS/CTS access mechanism
outperformed the basic access mechanism because
of collisions. In RTS/CTS access mechanism,
collisions occurred for control frames while in the
basic access collisions occurred for data packets as
well as control frames.
6 CONCLUSION
In this study extensive experiments were carried out
using ns2 simulation software to investigate the
performance of the IEEE 802.11 MAC protocol by
varying the number of active stations and varying
the channel data rate.
Both MAC protocol mechanisms, i.e. the basic
access and RTS/CTS access mechanisms were
employed. The effect of traffic types (i.e. CBR and
VBR) on the performance of the access mechanisms
was also analysed.
The study indicated that increasing the number
of active stations had an impact on the average
throughput when the basic access mechanism was
used.
High data rates improved the average
throughput, but degraded the channel utilisation.
This was because the control frames were sent at
low data rate (1 Mbps).
The basic access mechanism outperformed the
RTS/CTS access mechanism when the number of
active stations was small. For a large network size,
greater than 15 stations, the RTS/CTS access
mechanism outperformed the basic access
mechanism.
In the future a detailed evaluation of QoS
parameters for various applications such as audio,
video, file transfer and data will be carried out for
small and large networks. Furthermore, the network
parameters will be used for predicting the QoS and
other network conditions as an approach for
improving the protocol performance.
REFERENCES
Bianchi, G., 2000. Performance Analysis of the IEEE
802.11 Distributed Coordination Function. IEEE
journal on selected Area in Communication. vol. 18,
no. 3, Page(s): 535-547.
Broch J, Maltz, D., Johnson D., Hu, Y-C., 1998. A
performance Comparison of Multi-hop Wireless Ad-
hoc Network Routing Protocol. In proceedings of
IEEE/ACM MobiCom, Page(s):85 - 97.
Cali, F.; Conti, M.; Gregori, E., 2000. Dynamic Tuning of
the IEEE 802.11 Protocol to Achieve a Theoretical
Throughput Limit. IEEE/ACM Transactions. On
Networking, vol.8, no. 6, Page(s): 785-799.
Coverdate, P., 2000. ITU-T study group 12: Multimedia
QoS requirements from a user perspective. Workshop
(a): Data packet drop due
to buffer for CBR/basic.
(b): Data packet drop due
to buffer for VBR/basic.
(c): Data packet drop due to
b
uffer for CBR/RTS.
(d): Data packet drop due
to buffer for CBR/RTS
.
Figure 8: Percentage of buffer drop vs. Number of stations.
ICETE 2005 - WIRELESS COMMUNICATION SYSTEMS AND NETWORKS
34
on QoS and user perceived Transmission Quality in
Evolving Networks.
Haitao, W.; Yong P.; Keping, L.; Shiduan, Ch.; and Jian,
M., 2002. Performance of reliable transport protocol
over IEEE 802.11 wireless LAN: analysis and
enhancement. Proceedings IEEE INFOCOM 2002
Conference on Computer Communications. Twenty-
First Annual Joint Conference of the IEEE
Computer and Communications Societies, vol.2,
pt. 2, Page(s): 599-607.
IEEE standard 802.11, 1997. Wireless LAN Medium
Access Control (MAC) and Physical Layer (PHY)
Specifications.
IEEE standard 802.11, 1999. Wireless LAN Medium
Access Control (MAC) and Physical Layer (PHY)
Specifications.
Kleinrock L. and F. Tobagi A., 1975. Packet Switching in
Radio Channels: Part 2: The Hidden Terminal
Problem in Carrier Sense Multiple-Access and the
Busy-Tone Solution. IEEE Transaction on
Communication COM- 23, Page(s):1417–1433.
ns2, The network simulator-ns-2. [Online]. Last accessed
on 5 April 2005 at URL: http://www.isi.edu/nsnam/ns/
Sweet, C.; and Sidhu, D., 1999. Performance Analysis of
the IEEE 802.11 Wireless LAN Standard. [Online].
Last accessed on 30 August 2004 at URL:
http://www.csee.umbc.edu/~sweet/papers/
Thruong H, Vannuccini, G., 2003, Performance
Evaluation of the QoS Enhanced IEEE 802.11e MAC
Layer. Vehicular Technology Conference, 2003. VTC
2003-Spring. The 57th IEEE Semiannual,
vol 2, Page(s):940 - 944.
A COMPARATIVE STUDY OF IEEE 802.11 MAC ACCESS MECHANISMS FOR DIFFERENT TRAFFIC TYPES
35