Kr2518691873

Prof Rekha Patil , Ashwini.R.Malipatil / International Journal of Engineering Research and
Applications (IJERA) ISSN: 2248-9622 www.ijera.com
Vol. 2, Issue5, September- October 2012, pp.1869-1873
MAC Enhancement for Performance Improvement of Routing
Protocol
Prof Rekha Patil* , Ashwini.R.Malipatil**
*HOD & Guide Dept. of Computer Science and Engineering
Poojya Doddappa Appa Collage of Engineering Gulbarga 585103, India
**Dept. of Computer Science and Engineering
Poojya Doddappa Appa Collage of Engineering Gulbarga 585103, India

ABSTRACT
Mobile Ad hoc network allow wireless IEEE 802.11 standards support multiple data rates
devices to dynamically self organize into arbitrary with dynamic rate switching capability. IEEE 802.11
and temporary wireless topologies without fixed network, the diversity of data rate could lead to a
communication infrastructure. One of the most phenomenon known as performance anomaly. When
popular MAC layer protocol used in Ad hoc the host detects repeated unsuccessful frame
network is 802.11b. Since the IEEE 802.11 transmission, it decreases its bit rate from nominal
standard support multiple data rates at the 11Mbps to 5.5, 2 or 1Mbps. If ther is at least one host
physical layer, many rate adaptation mechanisms with a lower rate the IEEE 802.11 cell presents the
have been proposed to utilize this multi rate performance anomaly that is the throughput of all
capability by dynamically switching rates in other hosts transmitting at the higher rate is degraded
response to different channel conditions. Because below the level of the lower rate. To solve this
of varying channel condition, 802.11b products performance anomaly the idea of cross layer design
degrade the bit rate from the nominal 11Mbps to (CLD) is used in proposed work.
5.5,2 or 1Mbps. When a host detects repeated
unsuccessful frame transmission, it decreases its The whole idea behind CLD is to combine
bit rate. If there is at least one host with a lower the resources available in the different communities,
rate, 802.11 cells presents a performance anomaly. and create a network which can be highly adoptive by
The throughput of all the hosts transmitting at sharing state information between different processes
higher rate is degraded below the level of the or modules in the system. The interactions between
lower rate. the MAC and Routing layer is achieved by popular
In IEEE 802.11b networks, the guarantee routing protocol AODV. Nodes can acquire the
of equal transmission opportunity for all whole path’s rate information which can represent the
participating nodes results in performance whole path’s data transmission capability and uses it
anomaly. We propose a new cross layer MAC as criteria for allocating the channel.
mechanism, which allows nodes acquire the whole AODV routing algorithm is quit suitable for
path’s rate information by exchanging dynamic self organizing network. AODV provides a
information between MAC and Routing layer at loop free routes there by routes are created on
every node. The interactions between MAC and demand. There is no need of periodic routing
Routing Layer are fully exploited to achieve updates.
higher performance using AODV routing
protocol. Nodes acquire the whole path’s rate 2 RELATED WORKS
information by following a cross layer design Mobile Ad hoc networks (MANETs) are
approach. We evaluate the performance measured often characterized as dynamic wireless networks
on the network simulator OMNet++. that do not rely on any form of fixed supporting
infrastructure to operate. Conventionally,
Keywords: AODV, Cross layer, IEEE 802.11b, communication protocols and services in MANETs
MANET, Routing. are designed to adapt in response to traffic demands
and changes in network topology and link
1 INTRODUCTION characteristics [1] [18]. Many rate adaptation
Ad hoc networking is a concept in computer mechanisms have been proposed to improve the
communication, which means that the users wanting performance of multi rate MAC protocol. RBAR
to communicate with each other form a temporary (Receiver Based Auto Rate) as one of the most
network, without any form of centralized popular multi rate MAC protocol was proposed by
administration. Each node participating in the Gaven in [2]. Rate adaptation is the process of
network acts as both host and router and willing to dynamically switching data rates to match the
forward packet for other nodes. channel conditions, with the goal of selecting the rate
that will give the optimum throughput for the given

1869 | P a g e

channel conditions. The novelty of RBAR is that its architecture is cross layer design with respect to the
rate adaptation mechanism is in the receiver instead reference architecture.
of the sender. Compared with its poor channel quality
estimation and the dependence on RTS/CTS packet, AODV Routing protocol provides on
more problems of RBAR came from performance demand route discovery in mobile ad hic networks. In
anomaly. In [3], Martin Heusse et al, pointed out that [14] it uses an on demand approach for finding
the diversity of data rate in a wireless network could routes, that is, a route is established only when it is
lead to the performance anomaly in IEEE 802.11b required by a source node for transmission of data
networks, because 802.11 DCF make sure every packets. Power Aware AODV (PAAODV) protocol
node, whether high rate or low rate has the same for ad hoc networks [15] [19], which is an
opportunity to access the channel. Thus, low-rate enhancement of existing AODV ad hoc routing
nodes will occupy the channel much longer than protocol. The principal goal behind devising
high-rate nodes eventually leading to the PAAODV is to optimally reduce power consumption
performance degradation of the entire multi-rate to a possible minimum level in an ad hoc network
network. As pointed by Tan and Guttsg, the with the connectivity of the network not being
throughput-based fairness, which is widely accepted disrupted.
in single-rate network, could not suit multi-rate
network well. In [4], a new fairness notion known as 3 PROPOSED TECHNIQUE
time based fairness is considered. In time based In our work we develop a unique solution to
fairness, each node receives an equal share of the deduce the performance anomaly in MANET by
wireless channel occupancy time. It is demonstrated introducing a Cross layer MAC mechanism to
in [5] that the time based fairness can achieve higher enhance the performance. A cross layer MAC
aggregate throughput. mechanism for selection of route from source to
destination acquire the whole path’s rate information
Several solutions were proposed to solve and allocation the channel based on the relative ratio
performance anomaly. Initial contention window of competitive path’s data transmission capability.
adaptation [6] achieves the time based fairness by
adjusting its CWmin inversely proportional to the
data rate to make sure the high rate node have more
opportunity to access the channel. It was suggested in
[7] that performance anomaly can be reduced by
adjusting MAC frame sizes according to data rates,
which could be achieved by controlling the size of
maximum transmission unit(MTU). In [8], Sadeghi et
al. proposed a solution called OAR (Opportunistic
Auto Rate) protocol, which allows the nodes with
high data rate to transmit multiple packets in order to
take advantage of the high rate channel. Yongho
proposed a multi rate Aware Sub layer (MAS) [9]
which is independent of IP protocol to utilize the Fig 1 illustrate flow of data using cross layer
MAC and Physical layer when packets are routed in
MANETs. The key function of MAS is to change its We use “link” to refer to the point to point
next hop node to another node through which higher connections between two nodes through a wireless
data rate is available. Following the idea of MAS, MAC, where as we use a “path” to refer to end to end
many cooperative relay based medium access connections between two nodes through AODV
schemes are proposed to enable the full utilization of routing protocol. Two new sub structures are
the multi rate channel characteristics. RBMAC introduced to store its neighbor nodes MAC address
(Relay Based MAC) [10], Relay enable DCF [11] and data rate.
and CRBAR (Cooperative Relay Based Auto Rate
MAC) [12] all take the whole path’s data rate or the
path’s average end to end delay into consideration for
selection of appropriate path.

In recent years, the idea of cross layer
design (CLD) has drawn more and more attentions.
In [13] [16] [17], the cross layer survey says
protocols can be designed by violating the reference
architecture, by allowing direct communication
between protocols at nonadjacent layers or sharing
variable between layers. Such violation of the layered Fig 2 illustrates storage structure

1870 | P a g e

A new structure denoted by Path List, is introduced path rate transmits to the source node. Only a source
to store path’s information. The first two node of the path can update its Path List based on
substructures of Path List are used to store path’s information piggyback by RREP because only when
source node and destination node’s IP addresses. The receiving RREP, source node has acquired the whole
Path Rate substructure is used to record the reciprocal path’s rate information. Source node stores the path’s
of the time that spent transmitting unit data packet information in its Path List which includes source
from path’s source node to the destination node. Path node’s IP address, destination IP address, Path Rate
Rate, which can be recognized as the average data and Path Level. Hence the source node transmits the
rate of the whole path, represents the whole path’s data to the destination.
data transmission capacity so nodes on the same path
have the same Path Rate. Competitive node’s 3.1 METHODOLOGY
competitive level which is the ratio of its own Path Step 1: Generate RREQ packet at the source node.
Rate to the competitive node’s Path Rate, is stored in Each RREQ packet is appended with a new field path
Path Level substructure. rate.
Step 2: Each intermediate node transmits the RREQ
The source node generates RREQ packet for packet unless it finds itself as a destination node.
the route discovery from source to destination. Each Step 3: Each intermediate node takes its MAC
RREQ packet generated by the source node is QUEUE SIZE and calculates its RP value which is
appended with a new field of path rate. Source node added to the RREQ packet.
of a path needs to get the rate information of the Step 4: Once the RREQ packets are received by the
whole path. The interactions between MAC and destination node from different paths, the destination
Routing layers are achieved with the protocol calculates the cost of each path.
AODV, a widely used ad hoc routing protocol. Each Step 5: Destination compares the current cost with
node calculates its MAC QUEUE SIZE which is the calculated cost and selects the best path.
further used for the calculation of the rate parameter Step 6: Destination generates RREP packet for the
(RP): best path.
Step 7: Source node on receiving RREP packet
RP=11.2e+6/nbCount *MAC_QUEUE_SIZE+1 updates its Path List with the Path Rate and Path
(1) level and transmits the data

Where RP=Rate Parameter, nbCount= 4 SIMULATION
number of neighbor count, MAC_QUEUE_SIZE= The simulation is carried out in OMNet++
Queue size of each node. All the neighbor nodes 3.3 network simulator. The setup of the simulation
receiving RREQ packet adds its RP value to the rate depicts an ad hoc network that consists of a varying
field of the RREQ packet accept the destination node. number of Mobile Hosts (MHs) that move randomly
in a square field free of obstacles. OMNET++ is a
Number of RREQ packets reaching to freely distributed, object oriented, modular, discrete
destination contains the total RP value of all the event simulator written in C++. It is designed for
nodes lying as intermediate nodes in the path from general purpose discrete event simulation and
source to destination. Destination node stores all the provides model libraries for communication protocols
RREQ packets from different path in a route cache. and network systems. A Network Descriptor (NED)
The Path Rate which is average data rate of the whole language is used to model definitions based on
path is calculated at the destination which is system modules written in visual C++.
considered as the cost:
Simulation is carried out for the proposed
rp=rp+RP (2) technique in MANET with the routing layer along
with the comparison of present system. The
Cost=rp+en (3) simulation is carried out by varying parameters such
as nodes, sessions, area, pause time, packet rate, and
Where rp is the rate parameter of the whole path, simulation time. We measure the out put parameters
Cost is the sum of rate parameter and energy used. such as throughput (Mbps), packet delivery ratio (%),
control overhead (sec) and latency (sec).
cost=Cost/hopcount (4)

Where cost is the average cost for the whole path.

The destination node compares the current
cost with the calculated cost of each path and then
selects the best path. Once the path is been selected
the destination generates RREP packet with new field

1871 | P a g e

4.1 Simulation Results:
Pause time Vs Packet Delivery Ratio

120

Packet Delivery Ratio(%)
100

80
Proposed
60
Present
40
20

0
5 10 15 20 25 30 35 40 45 50
Pause time(sec)

Fig 6 illustrate Pause time Vs Packet delivery ratio
Fig 3 illustrates Throughput with the varying
Simulation time. Packet delivery ratio increases with the
increase in the pause time which indicates a complete
The above graph shows the performance of stability in the network and hence reduces packet
the proposed technique in terms of throughput. The drop. Hence the packet delivery ratio is best than the
throughput is much better than the present system. present.

5 CONCLUSION AND FUTURE WORK
The traditional techniques have some
limitations when solving performance anomaly
especially when dealing with the multi hop network
because in MANET the network condition and route
quality differs from time to time due to various issues
like node mobility, no power control, congestion in
the network and so on which ultimately affects the
reception quality and effective packet and data rate.

Therefore in our work we propose a novel
Fig 4 illustrate Simulation time Vs Packet delivery technique for QoS provisioning in MANET by
ratio calculating the route cost based on both energy and
Graph elaborates packet delivery ratio with Rate in a node. The cost is incorporated in RREQ.
respect to the varying simulation time. The proposed Destination forwards RREP only if cost of the most
system provides high packet delivery compared to the recent RREQ is better than the already ensured
present system. RREQ. Thus the work utilizes the advantages of
Paus e tim e Vs Throughput multipath routing without adding extra overhead
through route caching.
12

10
Results show that the system performs better
Throughput(Mbps)

8 with the cost based AODV. The system can be
Proposed
6 further improved by comprehensive Rate
Present
4
measurement technique which should measure rate
2
from Bandwidth, Delay and collision in the channel
and interference.
0
5 10 15 20 25 30 35 40 45 50
Paus e tim e (Se c) REFERENCES
[1] Sridhar Iyer,”Mobile Adhoc Networks”,IIT
Fig 5 illustrate Pause time Vs Throughput Bombay, CIT 2000.
[2] G.Holland, N.Vaidya,”A Rate Adaptation
Graph illustrates the throughput with respect MAC Protocol for Multi Hop Wireless
to varying pause time. The proposed technique shows Networks,” Proceedings of
high performance with the high throughput compared ACMMOBICOM’01,Rome,2001.
to the present. The throughput increases with the [3] M.Heusse, F.Rousseau, G.Berger-
increase in the pause time which decreases the high Sabbateland A.Duda,”Performance anomaly
mobility in MANET. of 802.11b,”Proc.IEEE
Int.conf.Comput.Commun.(INFOCOM),PP.
836-843,Apr.2003.

1872 | P a g e

[16] Philipp Hurni, ”Cross Layer Design in
[4] Tan, G., and GuRag, J.,”Time Based Wireless Networks” Universitat Bern 2008.
Fairness Imerformance in Multi rate [17] Frank Aune, “ Cross Layer Design Tutorial”
Wireless LANS,”Proceedings of Norwegain University of Science and
USENIX’04,USA, 2004. Technology, Dept of Electronics and
[5] A.V. Babu, and Lillykutty Telecommunications, Trondheim, Norway
Jacob,”Performance Analysis of IEEE 2004.
802.11 Multirate WLANs:Time Based [18] Jun-Zhao Sun, “Mobile Ad Hoc
Fairness Vs Throughput Based Networking: An Essential Technology for
Fairness,”International Conference on Pervasive Computing” Media Team,
Wireless Networks,Communications and Machine Vision and Media Processing Unit,
Mobile computing, 2005. Infotech University of Oulu, Finland 2001
[6] H.Kim, S.Yun, I.Kang, and [19] Liang Cheng, Anu G. Bourgeois, Bo Hyun
S.Bahk,”Resolving 802.11 performance Yu,” Power Management in Wireless Ad
anomalies through QoS Hoc Networks Using AODV” Department
differentiation,”IEEE Commun. Lett., of Computer Science, Georgia State
Vol.9,no.7,pp.655-657,jul.2005. University, Atlanta 2005.
[7] Yoo, S., Choi, J., and Hwan, J.,”Eliminating [20] Andra’s Varga, ”OMNet++ Discrete EEvent
the performance anomaly of Simulation System” User manual Version
802.11b,”LNCS,1055-1062,2005. 3.0, 2004.
[8] Sadeghi, B., Kanodia, V., Sabharwal,A., and
Knightly,E.,”Opportunistic Media Access
for Multi rate Ad Hoc
Networks,”MOBICOM,2002.
[9] Yongho S., Jaewoo P, and Yanghee
C.,”Multirate Aware Routing Protocol for
Mobile Ad Hoc Networks,”Proc.IEEE
Vehicular Technology Conf.,pp.22-
25,Apr.2003.
[10] Zeng W, Tan H, and Suda T,”A Relay
Based MAC Protocol to Support Multi rate
Feature in Mobile Ad Hoc Networks,”Proc.
Of The Second Annual International
Conference on Mobile and Ubiquitous
Systems: Networking and
Services,2005:145-154.
[11] Hao Zhu, and Guohong Cao,”A Relay
Enabled Medium Access Control Protocol
for Wireless Ad Hoc Networks” IEEE
Transations on Mobile Computing, Vol.5,
NO.9, Sep.2006,pp.1201-1214.
[12] Tao Guo, and Rolando Carrasco,”CRBAR:
Cooperative Relay Based Auto Rate MAC
for Multirate Wireless Networks” IEEE
Transation on Wireless Communication,
VOL.8, NO.12, Dec.2009,pp.5938-5947.
[13] Vineet Srivastava, Mehul Motani, “Cross
Layer Design: A Survey and the Road A
head,” Institute for Infocomm Research and
National University of Singapore 2005.
[14] Perking C, Royer E B, and Das S,”Ad Hoc
on demand distance vector (AODV)
routing,” RFC3561-2003.
[15] Binod Kumar Pattanayak, Manoj Kumar
Mishra, Alok Kumar Jagadev, Ajit Kumar
Nayak Journal of Convergence Information
Technoligy,” Power Aware Ad Hoc On
demand Distance Vector (PAAODV)
Routing for MANETS” Volume 6, Number
6, June 2011.

1873 | P a g e

Kr2518691873

More Related Content

What's hot

Viewers also liked

Similar to Kr2518691873

Kr2518691873