Channel Aware Mac Protocol for Maximizing Throughput and Fairness

International Journal of Research in Computer Science
eISSN 2249-8265 Volume 3 Issue 5 (2013) pp. 1-9
www.ijorcs.org, A Unit of White Globe Publications
doi: 10.7815/ijorcs. 35.2013.069
www.ijorcs.org
CHANNEL AWARE MAC PROTOCOL FOR
MAXIMIZING THROUGHPUT AND
FAIRNESS
1
G.Suseendran, 2
E.Chandrasekaran
1
Ph.D. Research Scholar, Department of Mathematics
Presidency College (Autonomous), Chennai, INDIA
Email: suseendranphd@gmail.com
2
Associate Professor, Department of Mathematics
Presidency College (Autonomous), Chennai, INDIA
Email: e_chandrasekaran@yahoo.com
Abstract: The proper channel utilization and the
queue length aware routing protocol is a challenging
task in MANET. To overcome this drawback we are
extending the previous work by improving the MAC
protocol to maximize the Throughput and Fairness. In
this work we are estimating the channel condition and
Contention for a channel aware packet scheduling
and the queue length is also calculated for the routing
protocol which is aware of the queue length. The
channel is scheduled based on the channel condition
and the routing is carried out by considering the
queue length. This queue length will provide a
measurement of traffic load at the mobile node itself.
Depending upon this load the node with the lesser
load will be selected for the routing; this will
effectively balance the load and improve the
throughput of the ad hoc network.
Keywords: Mobile ad hoc networks (MANETs), MAC
protocol, Throughput and Fairness.
I. INTRODUCTION
A. MANET
Mobile ad hoc networks (MANETs) are self-
organizing networks that provide an efficient solution
when centralized control is infeasible like emergency
and rescue operations, disaster-relief efforts, etc. A
key design objective in MANETs is to achieve high
network throughput while maintaining energy-
efficient wireless communications for mobile
terminals. To achieve this objective, efficient design
of the MAC layer is necessary in order to resolve
channel-contention and reduce packet collisions.
Several important issues like energy efficiency,
fairness, or quality of service (QoS) provision need to
be carefully considered when designing MAC
protocols for MANET. For the system’s quality of
service (QoS) requirements, a MAC protocol for ad
hoc networks shares the medium and the available
resources in a distributed manner, and allows for
efficient interference management. Despite all the
research done on MAC protocols thus far, only a
limited number of works have considered an
interference channel that is both stochastic and
continuous in time [1][2][3].
B. Fairness Issues in 802.11
The IEEE 802.11 defines two MAC protocols, i.e.,
Point Coordination Function (PCF) and Distributed
Coordination Function (DCF), but only DCF is used
in MANETs since PCF requires base stations.
DCF is a CSMA/CA based protocol, this protocol
has two main components: Carrier Sensing (CS) and
Collision Avoidance (CA). Since CS and CA work in
a distributed manner without having precise
information, they cause unfairness. The CA
algorithms and imprecise EIFS problem, the hidden-
terminal problem and signal capture, which are the
two most important issues in wireless networks, also
affect the MAC fairness. Unfairness could result from
different opportunities of channel access.
There are two major sources for unequal channel
access opportunities: the backoff mechanism and
location. While the backoff mechanism is broadly
used in MAC protocols for MANETs to reduce
collisions and achieve high channel efficiency, it
always favors the node that just successfully seized
the channel. As a result, different nodes may use
different backoff window, leading to different
transmission probabilities and consequently short-
term unfairness as well as long-term unfairness. To
achieve fairness among all nodes, the network’s
aggregate throughput, namely, efficiency often has to
be sacrificed. The MAC protocol and the routing the
overall performance of protocol interact with each

2 G.Suseendran, E.Chandrasekaran
www.ijorcs.org
other, which in turn affect a MANET, it is very
important to characterize the interaction between the
MAC fairness and the routing mechanism. Fairness in
wireless ad hoc networks has been studied under
various network scenarios. Many algorithms has been
proposed to achieve fairness among single-hop flows,
but they do not consider multi-hops flows, which
reflect the reality in wireless ad hoc networks
[3][4][5].
C. Problem Identification
In our previous paper, we have proposed an
interference reduction technique in MANET using
hidden markov model (HMM). Initially, node that
receives the RREQ packet calculates its received
signal power and compares with predefined threshold
values namely, Pmin and Pmax. Based on these
comparison initial transmission values are set. During
transmission, RTS message is sent using initial
transmission power. RTS includes interference value
predicted by the source. Nodes use hidden markov
model (HMM) to predict their interference value. On
receiving RTS, the destination calculates its
interference and sends it along with CTS to the source
with the power level Pini. While receiving the CTS,
the source calculates the minimum power required for
transmitting data using the interference value of the
destination and transmits the data with that power
value. Finally, the destination uses the interference
value of the source for transmitting the ACK message.
Drawbacks
− The proposed technique does not deal with
the proper channel utilization.
Estimation of queue length is not considered which
results in the throughput and fairness of the MAC
protocol.
II. RELATED WORK
Fan Wang et al., in paper [1] have proposed a
distributed, single-channel MAC protocol (GMAC)
that is inspired by game theory. In GMAC, each
transmitter computes a utility function that maximizes
the link’s achievable throughput. The utility function
includes a pricing factor that accounts for energy
consumption. GMAC allows multiple potential
transmitters to contend through an admission phase
that enables them to determine the transmission
powers that achieve the Nash equilibrium (NE). The
advantage of this approach is that GMAC
significantly improves the network throughput over
the 802.11 scheme and over another single-channel
power-controlled MAC protocol (POWMAC). These
gains are achieved at no extra energy cost.
P. Priakanth et al., in paper [4] have proposed a
topology-aware MAC protocol which attains fairness
across multi-hop flows and minimizes the energy. The
proposed protocol estimates the feasible bandwidth
and channel condition of each wireless link by
monitoring its traffic and calculates a combined score.
The score is sent along with the data packets of the
flows using any routing protocol. Then transmission
is allowed only for those nodes with high scores.
Nodes attempting to access the wireless medium with
a low score will be allowed to transmit again when
their score becomes high.
Zhifei Li et al., in paper [5] by using static routing
they have first studied the MAC fairness by
considering three factors including the hidden
terminal, capture, and imprecise EIFS, which are all
related to the signal attenuation property of the
wireless networks. Since the MAC protocol and the
routing protocol interact with each other, we then
study the interaction between MAC fairness and
routing mechanism by replacing static routing with
AODV and DSDV. The main lessons learnt from this
studies include: (i) should distinguish between the
MAC fairness and the overall system fairness, (ii) in
studying the MAC fairness, should isolate the factors
(e.g., routing protocol) that are not directly related to
the MAC layer, and (iii) to study the overall system
fairness, in addition to the study of MAC fairness, we
should also consider the effects of many other factors
such as the routing mechanism, traffic pattern,
mobility pattern etc.
Xuemei Gao et al., in paper [6] have proposed a
load-aware routing protocol using two load metrics
for route selection, which include MAC layer channel
contention information, and the number of packets in
the interface queue. MAC layer contention
information provides an accurate estimation of
neighbor nodes’ state, and queue length provides a
This load-aware routing protocol can effectively
balance the load and improve the performance of the
ad hoc network. Impacts of these load metrics on the
routing performance are studied.
Mun Choon Chan et al., in paper [7] they present,
CaSMA, a scheduling mechanism for mobile ad hoc
networks (MANETs) that takes into account both the
congestion state and end-to-end path duration. This
scheduling mechanism is termed Channel aware
scheduling for Mobile Ad hoc networks (CaSMA),
where the term channel-aware is used to indicate both
the congestion state and the end-to-end path duration.
CaSMA is complimentary to packet scheduling
scheme that utilizes only local channel information,
and can be added to these schemes.

Channel Aware Mac Protocol for Maximizing Throughput and Fairness 3
www.ijorcs.org
III. PROPOSED WORK
A. Overview
We are extending the previous work by improving
the MAC protocol to maximize the Throughput and
Fairness. In this work we are estimating the channel
condition and Contention for a channel aware packet
scheduling and the queue length is also calculated for
the routing protocol which is aware of the queue
length. The channel is scheduled based on the channel
condition and the routing is carried out by considering
the queue length. This queue length will provide a
Depending upon this load the node with the lesser
load will be selected for the routing; this will
effectively balance the load and improve the
throughput of the ad hoc network.
B. System Design
In our previous paper, we have proposed an
interference reduction technique in MANET using
hidden markov model (HMM). Initially, node that
receives the RREQ packet calculates its received
signal power and compares with predefined threshold
values namely, Pmin and Pmax. Based on these
comparison initial transmission values are set. During
transmission, RTS message is sent using initial
transmission power. RTS includes interference value
predicted by the source. Nodes use hidden markov
model (HMM) to predict their interference value. On
receiving RTS, the destination calculates its
interference and sends it along with CTS to the source
with the power level Pini. While receiving the CTS,
the source calculates the minimum power required for
transmitting data using the interference value of the
destination and transmits the data with that power
value. Finally, the destination uses the interference
value of the source for transmitting the ACK message.
We are extending this work by improving the
MAC protocol to maximize the Throughput and
Fairness. In this extension we are estimating the
channel condition and Contention for a channel aware
packet scheduling and the queue length is also
calculated for the routing protocol which is aware of
the queue length.
1. Channel Condition: In this approach the end to end
channel quality is represented in the form of path
lifetimes. The channel state keeps changing
continuously hence the end to end path will be valid
for a temporary period of time. The term path lifetime
is used to define the time interval for which the path
associated for a flow is valid or exists. Suppose the
lifetime of each and every link of path P from node i
to node j is estimated as 𝑙1, 𝑙2, … . , 𝑙 𝑛, then the path
lifetime is given by;
𝑃𝑗 = 𝑚𝑖𝑛(𝑙1, 𝑙2, … . , 𝑙 𝑛)
2. Global Ideal Scheduler: Now let us consider a
simple model with multiple flows over a single
bottleneck link where we have a single scheduler.
After the single shared link (with infinite lifetime),
these flows use different (non-shared) links with
different lifetime. Let 𝑆 𝑔 be the global scheduler,
which will schedule the flows (q flows). Let us
consider a single continuous period “p” of “q” flows,
with arrivals within this continuous period, and no
further arrivals.
Use of QS/RLT to Approximate Ideal Scheduler
The scheduling approach is considered where
queue with maximum value of
RLT
QS
is chosen,
where QS is queue size and RLT is the residual life
time. The
RLT
QS
acts as a request rate, hence serving
queue which has higher
RLT
QS
values first will result
in providing higher priority to flows which takes
short-lived paths.
Considering
RLT
QS
is not sufficient to provide
equal proportion of service, so we are considering a
simple model, which is a single snap-shot in time
where we have “n” flows with each flow “I” having
workload (number of packets) as iW . Let the service
time for all packets be 1 time unit. Now all the flows
have T (minimum packet inter-arrival time) and C
(maximum packet transmission time) set to 1, the
ii WP = i.e. RLT for each flow will be the same as
iW . Let the maximum number of packets the
scheduler serves in the given time duration (or the
maximum duration of time snapshot) be maximum of
iW values, termed as 𝑊𝑚𝑎𝑥 . Let the of packets served
for flow i be iX and use 𝑅 𝑠𝑢𝑚 to represent sum of all
isR .
We know that QS either decreases or remain the
same, and RLT strictly decreases. Therefore request
rate 





RLT
QS
can either remain the same or increase.
Using the above model and notations we can rewrite
request rate as
∑−
−
n ji
ii
XW
XW
(1)
Let α (0 < α < 1) be the proportion (percentage) of
of iW service that any flow “i” receives at any given
time within the considered time duration. The
important point to note here is that, there is no one-to-
one mapping between the request rate considered and
proportion of service received (α). That is, if a flow i

www.ijorcs.org
has greater request rate than the other flow j, then it
may not mean that amount of the service
(proportionately, α) received by the flow i is lesser
than j. When the '
iP s varies to a larger extent, the
proportionate amount of services received by flows
can also vary to a larger extent.
For a special case where '
iP s are same, if a flow
has received lesser proportion of service than the
other flow, then its request rate will always be higher
than the other flow. Under these conditions, it can be
shown that serving by
RLT
QS
, results in fair
distribution of service.
If the iP values vary then we should avoid the
condition where short-lived flows can receive
proportionately greater service. This is achieved by
having an additional parameter termed as eligible −
service, for each flow. This eligible − service for any
flow i is equivalent to,
sumW
Wmax and is computed by
considering the sRi , which is given as follows:






∗∗
∗






∗
∑ =
max
max
max
1
T
C
W
T
C
W
T
C
W
n
j
i
i
j
i
i
i
(2)
Cmax and Tmax indicate the maximum possible
values of C and T, respectively. The first term
indicates the ratio of the work to be performed for a
flow i and the total amount of work considering all
flows. Whereas, the second term indicates the
maximum work that can be done, and this term, in
practice, is related to the maximum wireless link rate.
We update the eligible − service parameter only when
new flows arrive or existing flows leave. The priority
is given to flows by considering both the request rate
and eligible service. Higher priority is given to flows
whose request rate is high, and which has not yet
received its eligible-service. This parameter will
ensure that flows do not receive greater service (in
proportion) at the cost of other flows.
3. Schedulability: Now we will enhance the
approximation of ideal scheduler by considering end-
to-end packet scheduling. A set of flows Γ is said to
be “schedulable” (S) if none of the flows has packets
queued in the intermediate nodes at the end of their
respective continuous periods. Any set of flows at a
node that are schedulable over a link is termed as
“schedulable set”.
For the flow schedulability the following two cases
are considered.
Case 1: We have to consider that given a set of n
flows Γ = (Ti, Ci, oi), i = 1, 2, . . .n, how many of them
(m, m ≤ n) are schedulable over a link? (Schedulable
set).
First, let us begin with the schedulable set (ζ). A
schedulable set is derived as given below. Let us
assume that a node has n flows, of which it has to
choose m flows to form a schedulable set. The
necessary condition for a set of flows to be
schedulable over a link is given as
1
1
≤





∑
=
m
i i
i
T
C
(3)
The above necessary condition in rewritten interms
of the packets scheduled. Since the minimum packet
interarrival time of a flow i is Ti, there are at most (oi)
Ti packets arrived over channel i during the interval,
and which need at most (oi) Ti Ci units of time to
transmit. Now the summation of this time for all the
m flows should be less than the rmax, which is written
as
max
1
W
T
Pm
i i
i ≤







∑
=
(4)
As there are different combinations that are
possible in choosing m flows out of n flows (Cm n ).
The value of m is dependent on the Ci and Ti values.
For example, value of m becomes smaller for smaller
values of Ti. Hence, we have to decide on a specific
way to choose m flows out of n flows. We choose the
m flows considering the residual lifetime values of the
flows. Scheduling based on residual lifetime is similar
to earliest deadline scheduling (EDF). We sort all the
n flows in terms of the increasing residual lifetime,
and from this sorted set we choose the first m flows.
These m flows from our schedulable set ζ.
Case 2: Suppose there are n flows Γ = (Ti, Ci, oi), i =
1, 2, . . .n, of which m flows form a schedulable set ζ.
Now, given a new flow j, what is the maximum value
of its continuous period (oj), such that the new flow
will be subset of the schedulable set (may result in
preemption of a flow existing in the current
schedulable set).
If a node has a set of flows Γ passing through it,
we define a schedulable set ζ (ζ ⊆Γ) where ζ is the
set of flows which are schedulable at that particular
node. Let the maximum continuous period in the set ζ
be oj of some flow j. The schedulable set ζ also
satisfies the necessary condition provided above. Now
the maximum value of continuous period for a new
flow, say k to be schedulable is to be lesser than oj ,
and the arrival rate is lesser than or equal to j´s arrival
rate. That is, a new flow k with continuous period ok
will be schedulable, iff ok < oj and Ck Tk ≤ Cj Tj . This

www.ijorcs.org
is because; the schedulable set is built considering
two conditions - residual lifetime and the necessary
condition as equation 3.
If the continuous period of the new flow (k) is
lesser than the continuous period of a flow (j), where
flow j is both a member of the existing schedulable
set and has a maximum continuous period in the set,
then the new flow (k) will be added into the
schedulable set at the expense of this existing flow ( j,
which had maximum continuous period will be
preempted). In addition, the second condition (Ck Tk ≤
Cj Tj ) is important to make sure that the new
schedulable set does not violate the condition given
by the equation 3. Therefore, for a flow to become
eligible as a member of the existing schedulable set is
that its continuous period be lesser than the maximum
continuous period in the existing schedulable set.
The solution to the second case leads to the notion
of a flow i being “schedulable” (S) at node l. This
notion provides an important parameter in our
analysis, as it is used in two ways: (1) An end-to-end
measure of this value during the path set-up helps the
source to decide on initiating the traffic (2)
Intermediate nodes make their scheduling decision
based on these values, which can be updated by the
downlink neighbors whenever value changes. We
know that if a flow is schedulable at all the
intermediate nodes, then it is schedulable over the
path. The idea is analogous to the series of traffic
lights. It is useful to turn the first light green when all
the remaining lights will turn green within some
acceptable duration. This technique helps in
increasing the merit of a scheduler, as priorities are
given to packets which will be “completely served”.
The notion of schedulability takes on only binary
values (TRUE/FALSE). When we use this parameter
in the algorithm, the mechanism just makes the
decision for given values and existing conditions.
This decision process is used to build the schedulable-
list message, as described below, in the following
example.
Figure 1: Example for Scheduling Process
Let us consider the example in which three nodes
S, I, and D as shown in Figure 1. We will focus on a
single flow ‘a’ starting at node ‘S’, with intermediate
node ‘I’ and terminating at node ‘D’. Let {a, b, c, d}
be the flows at ‘D’. Let {2, 4, 4, 6} and {1/2, 1/4, 1/4,
1/6} be their continuous periods and rates (C T ),
respectively. Node ‘D’ chooses flows {a,b,c} as
schedulable following the condition given by equation
3, and creates a schedulability list message, which is
transmitted to the upstream neighboring nodes. When
‘I’ receives this message, marks flow ‘a’ as
schedulable at downstream (sets the schedulability
value to TRUE), and builds its own schedulable-list
(let it be {a, y, z}) and transmits it to its upstream
neighbors. In this manner, the schedulable-list
message flows upstream until it reaches source node
‘S’, which upon receiving will mark flow ‘a’ as
schedulable. If either the destination node or any of
the intermediate nodes does not include flow ‘a’ in
their schedulable list message, then the source node
will not set flow ‘a’ as schedulable.
C. Algorithm
while(a set of high-priority real-time queues)
{
if (q.schedulability = TRUE)
{
select the set of queues, such that for every
queue q,
}
else
{
select all the queues.
}
if (value QS(q) RLT(q) is the maximum and not
yet received eligible-service)
{
select queue q
}
if (tie)
{
select flow that has received least throughput
}
D. Contention and Queue Length Aware Routing
Initially the node “i” stays in IDLE state. When
packet arrives to node i, it will enter Packet-Arrival
state. In this state, if node i senses medium busy in
SIFS period, it identifies the channel is busy and
enters the Back-off state right away, and then change
the value of contention window. The station calculates
a random value called back-off time. The random
number resulting from the binary exponential back-off
algorithm is uniformly distributed in a range, called
the contention window, the size of this doubles each
time the attempt to transmit is deferred, until a
maximum size is reached for the range. If the medium
is idle for more than Distributed Inter Frame Space

www.ijorcs.org
(DIFS) then it will enter the Attempt state and delay a
random back-off time interval before the
transmission. The back-off timer is periodically
decremented by one for every time slot the medium
remains idle after the channel has been detected idle
for a period greater than DIFS. As soon as the back-
off timer expires, the station can access the medium.
In the Attempt state the node i will first issue the RTS
control packet and then waits for the CTS packet to
make sure the transmission is successful. If no CTS is
detected within a slot-time, node i will return Back-
off state immediately and double the value of CW else
it will transmit the data and wait for ACK. If no ACK
is detected within a slot-time, node i will return to
Back-off state immediately. Once the ACK is detected
then it means a transmission is successfully
transmitted, CW is reduced to its minimum value for
next transmission. The CW indicates that the medium
is busy and can be considered as a useful metric for
contention and traffic situation around this node. This
can replicate the contention of the channel roundly
and effectively. To mitigate the effect of traffic bursts,
CW is used to calculate the traffic load over a long
period. The calculation of the CW of a node is
performed every T seconds. The CW is calculated by
the below equation
sampleold CWCWaCW +∗= (5)
We set α to 0.3 to better reflect the current
condition of a node, which grants a higher priority to
the current sample. The number of packets in the
queue is a metric reflecting the traffic load of this very
mobile node. A mobile node with more traffic flows
passing through it usually has more packets in its
interface queue. Average queue size can indicate this
node’s traffic load in a long term. The calculation of
the average queue size is updated every T seconds
according to the following formula,
( ) sampleold qlengthqlengthqlength ∗−+∗= ββ 1 (6)
Where qlength denotes the average queue length
and sampleqlength denotes the current queue
length, β is constant. It is noteworthy that α and β can
be any number selected from the range [0, 1]. Based
on the CW and queue length, the local load of node i
can be calculated as follow,
( )
maxmax
1
qlength
qlength
CW
CW
Li ∗−+∗= γγ (7)
The selection of constant 𝛶 is to balance the
effects of the two factors. The small qlength means
the low load, the small CW reflects the benign channel
and these cause lower 𝐿𝑖. Therefore, the higher value
of the Li means the larger load.
1. Route discovery and maintenance: This routing
scheme is based on Dynamic Source Routing (DSR)
protocol with the following modifications:
(1) During the route discovery process, the
intermediate nodes are not allowed to send back route
replies even if they have routes to the destination in
their caches. The purpose is to get the up-to-date load
information along the whole route for each route
discovery.
(2) Suppose if there are multiple routes available,
route selection is based on the comprehensive route
load value defined in the previous section. When a
source node has packet to send and there is no
available route in its route cache, a route request
packet is initiated and flooded through the network.
Each node receiving this request will process and
forward it until it reaches the final destination. When
the intermediate node processes the request, it adds its
load value to the request message. Once the request
packet reaches the destination, the destination
generates a route reply packet and sends it back to the
source node. And the path is chosen with smaller
load. When some link on an active route is broken, the
source node is notified by a route error packet. The
source updates its route cache by removing any route
using this broken link, and then initiates another route
discovery process to find a new least load route if
necessary.
E. Advantages
− The fairness mechanism is added in the MAC
protocol to the interference reduction technique.
− This technique will increase the throughput
further and achieve fair channel utilization.
− The proposed technique consists of estimating
channel condition and queue level based on which
the channel is fairly allocated to all the nodes.
IV. SIMULATION RESULTS
A. Simulation Parameters
We evaluate our Channel Aware MAC protocol
(CAMAC) through NS-2 [8]. We use a bounded
region of 1000 x 1000 sqm, in which we place nodes
using a uniform distribution. The number of nodes is
100. We assign the power levels of the nodes such
that the transmission range of the nodes is 250 meters.
In our simulation, the channel capacity of mobile
hosts is set to the same value: 2 Mbps. We use the
distributed coordination function (DCF) of IEEE
802.11 for wireless LANs as the MAC layer protocol.
The simulated traffic is Constant Bit Rate (CBR). The
simulation topology is given in Figure 2.

www.ijorcs.org
Figure 2: Simulation Topology
The following table summarizes the simulation
parameters used
Table 1. Simulation Parameters
No. of Nodes 100.
Area Size 1000 X 1000
Mac 802.11
Simulation Time 50 sec
Traffic Source CBR
Packet Size 500
Transmit Power 0.660 w
Receiving Power 0.395 w
Idle Power 0.035 w
Initial Energy 10.3 J
Transmission
Range
250m
Routing Protocol AODV
Flows 2, 4, 6 and 8.
Error Rate 0.01 to 0.05
B. Performance Metrics
We compare the performance of our proposed
CAMAC with Channel-aware Scheduling mechanism
for MANETS (CaSMA) technique [7]. We evaluate
mainly the performance according to the following
metrics:
Received Bandwidth: It is the number of bits
transmitted to the destination through the channel.
Packet Lost: It is the number of packets dropped
during the transmission.
Delay: It is the amount of time taken by the packets
to reach the destination.
C. Results
1. Based on Flows
In our experiment we vary the number of CBR
traffic flows as 2, 4, 6 and 8 with 50 nodes.
Figure 3: Flows vs Received Bandwidth

www.ijorcs.org
Figure 4: Flows vs Fairness
Figure 5: Flows vs Delay
Figure 6: Flows vs packet Lost

www.ijorcs.org
From figure 3, we can see that the received
bandwidth of our proposed CAMAC is higher than
the existing CASMA technique.
From figure 4, we can see that the fairness of our
proposed CAMAC is higher than the existing
CASMA technique.
From figure 5, we can see that the delay of our
proposed CAMAC is less than the existing CASMA
technique.
Form figure 6, we can see that the Packet lost ratio
of our proposed CAMAC is less than the existing
CASMA technique.
2. Based on Error Rate
In our second experiment we vary the error rate as
0.01, 0.02, 0.03, 0.04 and 0.05 for 100 nodes.
Figure 7: Error Rate vs Received Bandwidth
Figure 8: Error Rate vs Fairness

www.ijorcs.org
Figure 9: Error Rate vs Delay
Figure 10: Error Rate vs Packet Lost
From figure 7, we can see that the received
bandwidth of our proposed CAMAC is higher than
the existing CASMA technique.
From figure 8, we can see that the fairness of our
proposed CAMAC is higher than the existing
CASMA technique.
From figure 9, we can see that the delay of our
proposed CAMAC is less than the existing CASMA
technique.
Form figure 10, we can see that the Packet lost
ratio of our proposed CAMAC is less than the
existing CASMA technique.
V. CONCLUSION
In this paper we are extending the previous work
by improving the MAC protocol to maximize the
Throughput and Fairness. In this work we are
estimating the channel condition and Contention for a
channel aware packet scheduling and the queue length
is also calculated for the routing protocol which is
aware of the queue length. The channel is scheduled
based on the channel condition and the routing is
carried out by considering the queue length. This
queue length will provide a measurement of traffic
load at the mobile node itself. Depending upon this
load the node with the lesser load will be selected for
the routing; this will effectively balance the load and
improve the throughput of the ad hoc network. The
advantage of this work is that, the fairness mechanism

www.ijorcs.org
is added in the MAC protocol to the interference
reduction technique. Future work concentrates on
developing a rate adaptation algorithm based on the
channel condition.
VI. REFERENCES
[1] Fan Wang, Ossama Younis and Marwan Krunz,
“Throughput-oriented MAC for mobile ad hoc
networks: A game-theoretic approach”, 2008 Published
by Elsevier B.V. doi: 10.1016/j.adhoc.2007.12.002
[2] Mariam Kaynia, Nihar Jindal and Geir E. Øien,
“Improving the Performance of Wireless Ad Hoc
Networks Through MAC Layer Design”, IEEE
Transactions on Wireless Communications, vol. 10, no.
1, January 2011, doi:
10.1109/TWC.2010.110310.100316
[3] Hongqiang Zhai, Jianfeng Wang, Xiang Chen, and
Yuguang Fang, “Medium Access Control in Mobile Ad
Hoc Networks: Challenges and Solutions”, 2006. doi:
10.1002/wcm.376
[4] P.Priakanth, Dr.P.Thangaraj, “Energy and Channel
Aware MAC Protocol to Achieve Fairness in Multi-Hop
Mobile Adhoc Networks”, IJCSNS International
Journal of Computer Science and Network Security,
VOL.8 No.4, April 2008.
[5] Zhifei Li, Sukumar Nandi, and Anil K. Gupta, “Study of
IEEE 802.11 Fairness and its Interaction with Routing
Mechanism”, In the Fifth IFIP International Conference
on Mobile and Wireless Communications Networks
(MWCN), Singapore, Oct. 27-29, 2003.
[6] Xuemei Gao, Xinming Zhang, Dong Shi, Fengfu Zou,
Wenbo Zhu, “Contention and Queue-aware Routing
Protocol for Mobile Ad h oc Networks”, IEEE
International Conference on Wireless Communications,
Metwork and Mobile Computing, 2007. doi:
10.1109/WICOM.2007.410
[7] Sridhar K N, Mun Choon Chan, “Channel-aware Packet
Scheduling for MANETs”, IEEE International
Symposium on a World of Wireless, Mobile and
Multimedia Networks, 2008. doi:
10.1109/WOWMOM.2008.4594864
[8] Network Simulator: http: ///www.isi.edu/nsnam/ns
How to cite
G.Suseendran, E.Chandrasekaran, " Channel Aware Mac Protocol for Maximizing Throughput and Fairness ".
International Journal of Research in Computer Science, 3 (5): pp. 1-11, September 2013. doi: 10.7815/ijorcs.
35.2013.069

Channel Aware Mac Protocol for Maximizing Throughput and Fairness

More Related Content

What's hot (16)

Similar to Channel Aware Mac Protocol for Maximizing Throughput and Fairness (20)

More from IJORCS (20)

Recently uploaded (20)

Channel Aware Mac Protocol for Maximizing Throughput and Fairness