Performance Evaluation of MANET Routing Protocols under CBR

Document Sample
Performance Evaluation of MANET Routing Protocols under CBR Powered By Docstoc
					                                                                                                                  ISSN:2229-6093
     M L Sharma,Noor Fatima Rizvi,Nipun Sharma,Anu Malhan,Swati Sharma, Int. J. Comp. Tech. Appl., Vol 2 (3), 392-400




 Performance Evaluation of MANET Routing Protocols under CBR and FTP
                             traffic classes
                               M.L Sharma madansharma.20@gmail.com
                               Noor Fatima Rizvi noorieazeem@gmail.com
                                Nipun Sharma nipunharitash@gmail.com
                                  Anu Malhan anumalhan@gmail.com
                                Swati Sharma swatiharitash@gmail.com


Abstract
Understanding the performance of routing protocols in ad hoc networks is a key feature to determine
which routing protocol is best suited for which type of network scenario. From the literature survey it was
found that there is a lot of work done on evaluating the performance [3] of various MANET routing
protocols for CBR traffic but there is very little work done for variable bit rate like FTP, TELNET type of
traffic.

So, in this paper it is proposed to evaluate and analyze the performance of proactive (WRP) and reactive
(AODV, DSR) routing protocols based on traffic generators like FTP under different network scenarios
like pause time, offered load (i.e. number of source destination pairs), node speed.

Keywords: Ad-Hoc Networks, Performance Evaluation, Packet Delivery Ratio, CBR, FTP



    1. Simulation Methodology
Simulation studies have been carried out using GloMoSim [1] network simulator. The modules have been
developed using VC++ programming language. GloMoSim is a scalable simulation library for wireless
network systems built using the PARSEC simulation environment GloMoSim is designed using a layered
approach similar to the OSI seven layer network architecture. Simple APIs are defined between different
simulation layers. This allows the rapid integration of models developed at different layers by different
people.

    2. Performance Metric
In this paper we have worked on Packet Delivery Ratio as the performance metric to evaluate and analyze
the performance of various routing protocols.

•       Packet Delivery Ratio
Packet delivery ratio is defined as the ratio of data packets received by the destinations to those generated
by the sources. This performance metric gives us an idea of how well the protocol is performing in terms
of packet delivery at different speeds using different traffic models. Mathematically, we can define as,




                                                                                                                        392
                                                                                                                                                                                                                  ISSN:2229-6093
     M L Sharma,Noor Fatima Rizvi,Nipun Sharma,Anu Malhan,Swati Sharma, Int. J. Comp. Tech. Appl., Vol 2 (3), 392-400




                                       ������������������������ ���������������� �������������������������������� �������������������������������������������������������� ���������������������������������������������������������������� ���������������� ������������������������ℎ ����������������������������������������������������������������������������������������
                             ��������

                            �
PDR (%) =

                                       ������������������������ ���������������� �������������������������������� �������������������������������������������������������� ������������������������������������������������������������������������ ���������������� ������������������������ℎ ������������������������������������������������
                            ��������=1

                                                                                                                       m

where,                  i, indicates the number of output file

                        m, indicates the total number of output files



    3. Simulation Environment
Our simulation considered a network of 50 wireless nodes placed randomly within a 1000 x 1000 m2 area
and transmission range of each node is 250 meters. CBR, FTP and TELNET data sessions are chosen.
Only a specified number of nodes out 50 will be engaged in data transfer which we specify as offered
load. This is done to see the impact of varied load on various performance metrics. However, during this
data transfer process all of the 50 nodes will operate in the background for providing necessary support
(i.e. routing/forwarding) to the ongoing communication in the network. The rate for CBR traffic is 2
Mbps while the data packet size is 512 bytes. For FTP and TELNET traffic the data rate and packet size is
chosen randomly by the simulator which uses a random number generator function to randomly select the
number of items to be sent, the size of each item and the size of control packets.

Each simulation is executed for 30 minutes. However, data packets are generated by the sources only
during last 800 seconds of simulation time. The initial transient problem illustrates that it requires some
time period for nodes to settle down in a MANET and then data transfer actually starts and it is shown
that the random waypoint model with zero minimum speed cannot reach a steady state over the course of
a simulation, and the level of mobility continuously decreases. This causes the metrics observed to
continuously decrease as well. Because of this, time average of these metrics cannot be reliably compared
by varying the maximum speed. As we have seen, even under the same maximum speed depending on
how long the simulation is run, the resulting average can drastically differ. To avoid initial transient
problem and the problem with Random Waypoint Mobility Model RWMM [2], we discard the initial
1000 seconds of simulation period and do not take minimum speed as 0s. Five runs with different seeds
have been conducted for each scenario and collected data is averaged over these runs. A summary of
salient simulation parameters are presented in Table 3.1.

                                     Table 3.1 Salient Simulation Parameters
            Parameter                                                                                                                                                 Value

Simulation Time                                            30 min (1800 sec)

Terrain Area                                               1000×1000 m2

Number of Nodes                                            50

Node Placement Strategy                                    Random

Propagation Model                                          Two-Ray Model

Transmission Range of each Node                            250 m




                                                                                                                                                                                                                     393
                                                                                                                    ISSN:2229-6093
       M L Sharma,Noor Fatima Rizvi,Nipun Sharma,Anu Malhan,Swati Sharma, Int. J. Comp. Tech. Appl., Vol 2 (3), 392-400




Mobility Model                               Random-Waypoint

Radio Type                                   Accumulated Noise Model

Network Protocol                             IP

MAC Protocol                                 IEEE 802.11 DCF

Routing Protocols                            AODV, DSR,WRP


      4. Network Scenarios
In the ad hoc network, we have simulated the following 3 different scenarios:

      (a) Pause Time
      (b) Offered Load (number of source destination pairs)
      (c) Node Speed

In each of the scenario, unless otherwise specified, simulation settings are same as shown in Table 4.1.

(a)       Pause Time
Pause time refers to the rest time of the node. The RWMM includes pause times between changes in
direction and/or speed. A node begins by staying in one location for a certain period of time (i.e. a pause
time). Once this time expires, the node chooses a random destination in the simulation area and a speed
that is uniformly distributed between [MIN SPEED, MAX SPEED]. The node then travels towards the
newly chosen destination at the selected speed. Upon arrival the node pauses for a specified time period
before starting the process again. In our simulation, we considered 10 m/s as an average node speed, 10
SDPs as offered load, random waypoint as mobility model and 0,500,1000,1500,1800 seconds as pause
time. Where, 0s pause time represent the continuous node mobility and 1800s pause time represents static
network environment.

(b)        Offered Load (Number of SDPs)
Offered load refers to the number of source destination pairs engaged in data transfer. For example, with
10 SDPs amongst 50 nodes, 10 source nodes and 10 destination nodes (i.e. 20 nodes in total) will be
engaged in data transfer. However, during this data transfer process, all of the 50 nodes (including the
above 20 nodes) will operate in the background for providing necessary support (i.e. routing/forwarding)
to the ongoing communication process in the network. In our simulation we considered 10 m/s as an
average speed and 0s pause time with offered load (i.e. number of SDPs) varied as 10,20,30,40 pairs.

(c)       Node Speed
Node speed refers to the average speed with which nodes move in the simulation area. We have used
random waypoint mobility model (RWMM), as it is widely used in MANET simulations [23]. In
RWMM, nodes move at a speed uniformly distributed in [MIN SPEED, MAX SPEED]. In our
simulation, we have considered 10 SDPs for data transfer and average node speeds considered are 5, 10,
15, 20, 25 m/s. Each node begins the simulation by moving towards a randomly chosen destination.
Whenever a node chooses a destination, it rests for a pause time. It then chooses a new destination and
moves towards the same. This process is repeated until the end of the simulation time. In this scenario,
however, pause time is set at 0s (i.e. nodes move continuously throughout the simulation period). This is




                                                                                                                          394
                                                                                                                                          ISSN:2229-6093
                             M L Sharma,Noor Fatima Rizvi,Nipun Sharma,Anu Malhan,Swati Sharma, Int. J. Comp. Tech. Appl., Vol 2 (3), 392-400




done to study the impact of continuous node mobility (i.e. worst case scenario) on the network
performance.

In this section, we present the results obtained via simulations followed by analysis. Packet delivery ratio,
average end-to-end delay, throughput, routing message overhead are the metrics used to evaluate and
analyze the performance of reactive (AODV,DSR) and proactive (WRP) routing protocols under different
types of traffic like CBR, FTP, TELNET.

                            5. Results
CBR traffic
Simulation Results
In this section, the simulation results under traffic type CBR for 3 different scenarios, namely, pause time,
offered load, node speed are shown.

                            A. Pause Time Scenario
Packet Delivery Ratio

                            100
                             90
                             80
Packet Delivery Ratio (%)




                             70
                             60
                             50
                             40                                                                                                AODV
                             30                                                                                                DSR
                             20                                                                                                WRP
                             10
                              0
                                  0                 500              1000                 1500                1800
                                                            Pause Time (in seconds)

                                        Figure 5.1 Impact of pause time on packet delivery ratio (CBR traffic)

In Figure 5.1 we observe the impact of pause time on packet delivery ratio. The results show that the
packet delivery ratio is maximum when the pause time is equal to the simulation time (i.e. when the nodes
in the network are static). The reactive protocol AODV shows the best performance with 99% packet
delivery at 1800s pause time. DSR has approximately 30% less delivery ratio than AODV when pause
time is less but in a static environment DSR has comparable packet delivery ratio. The WRP being
proactive protocol has lesser packet delivery ratio i.e. approximately 60% less delivery ratio than AODV
and 20% less than DSR. But in a static network it has a packet delivery ratio of around 73% which shows
an increase of more than 30% for pause time 0s with maximum mobility.




                                                                                                                                                395
                                                                                                                                           ISSN:2229-6093
                              M L Sharma,Noor Fatima Rizvi,Nipun Sharma,Anu Malhan,Swati Sharma, Int. J. Comp. Tech. Appl., Vol 2 (3), 392-400




B. Offered Load (Number of Source Destination pairs) Scenario

Packet Delivery Ratio

                               90
                               80
  Packet Delivery Ratio (%)




                               70
                               60
                               50
                               40
                               30                                                                                       AODV
                               20                                                                                       DSR
                               10                                                                                       WRP
                                0
                                    10                    20                      30                      40
                                                         Offered Load (No. of SDPs)


                                         Figure 5.2 Impact of offered load on packet delivery ratio (CBR traffic)

Figure 5.2 shows the impact of offered load (i.e. number of source destination pairs) on the packet
delivery ratio in a network of 50 nodes randomly placed with 0s pause time. The results show that for
reactive protocols, AODV and DSR, the delivery ratio degrades with increase in load. The AODV, having
a delivery ratio of more than 80% at load of 10 SDPs degrades to less than 70% at load of 40 SDPs. The
DSR with delivery ratio of 55% at 10 SDP load degrades to 30% at a load of 40 SDPs. While proactive
protocol WRP having the least delivery ratio among the three protocols there is a slight increase in the
delivery ratio with load. It has a delivery ratio of 35% at a load of 10 SDPs which increases to more than
40% at a load of 40 SDPs.




                                                                                                                                                 396
                                                                                                                       ISSN:2229-6093
          M L Sharma,Noor Fatima Rizvi,Nipun Sharma,Anu Malhan,Swati Sharma, Int. J. Comp. Tech. Appl., Vol 2 (3), 392-400




C. Node Speed Scenario

(a) Packet Delivery Ratio

                              100
                               90
                               80
  Packet Delivery Ratio (%)




                               70
                               60
                               50
                               40
                               30                                                                          AODV
                               20                                                                          DSR
                               10                                                                          WRP
                                0
                                    5             10            15             20          25
                                                       Node Speed (in m/sec)


                                        Figure 5.3 Impact of node speed on packet delivery ratio (CBR traffic)

The Figure 5.3 shows the impact of changing the speed, with which nodes move in an ad hoc network, on
the packet delivery ratio. In general, packet delivery ratio decreases with increase in average node speed.
The packet delivery ratio for AODV is approximately 100% which remains almost same for all node
speed. The DSR shows a decrease of 20% in delivery ratio when the average node speed increases from 5
m/s to 25 m/s. The packet delivery ratio for WRP decreases by 13% with increase in node speed. This is
because higher speeds cause frequent link changes and connection failures.

FTP Traffic
In this section, the simulation results under traffic type FTP for 3 different scenarios, namely, pause time,
offered load, node speed are shown.




                                                                                                                             397
                                                                                                                                    ISSN:2229-6093
                       M L Sharma,Noor Fatima Rizvi,Nipun Sharma,Anu Malhan,Swati Sharma, Int. J. Comp. Tech. Appl., Vol 2 (3), 392-400



A. Pause Time Scenario

Packet Delivery Ratio

                              100
                              90
                              80
  Packet Delivery Ratio (%)




                              70
                              60
                              50
                              40
                              30                                                                                         AODV
                              20                                                                                         DSR
                              10                                                                                         WRP
                                0
                                    0               500             1000              1500               1800
                                                           Pause Time (in Seconds)


                                         Figure 5.4 Impact of pause time on packet delivery ratio (FTP traffic)

In Figure 5.4, the impact of pause time on packet delivery ratio, for FTP traffic, is shown. The delivery
ratio for reactive protocols, namely AODV and DSR is approximately 100%. Both the protocols have
almost same delivery ratio at all pause times hence these show the independence of delivery ratio on
pause time. On the other hand the delivery ratio for proactive protocol WRP first decreases with increase
in pause time and then increases with pause time. The delivery ratio at 0s pause time for WRP is
approximately 100% which then decreases to 40% at pause time 1000s and again increases to 80% at
pause time 1800s (i.e. when network becomes static).

            B. Offered Load (Number of source destination pairs)

Packet Delivery Ratio

                              100
                               99
  Packet Delivery Ratio (%)




                               98
                               97
                               96
                               95
                               94
                               93                                                                           AODV
                               92                                                                           DSR
                               91                                                                           WRP
                               90
                                    10                 20                  30                  40
                                                     Offered Load (No. of SDPs)

                                         Figure 5.5 Impact of offered load on packet delivery ratio (FTP traffic)



                                                                                                                                          398
                                                                                                                                    ISSN:2229-6093
                       M L Sharma,Noor Fatima Rizvi,Nipun Sharma,Anu Malhan,Swati Sharma, Int. J. Comp. Tech. Appl., Vol 2 (3), 392-400




Figure 5.5 shows the impact of offered load (i.e. increasing number of source destination pairs) on packet
delivery ratio under FTP traffic. Results show that for all protocols packet delivery ratio is independent of
offered load. All the three protocols have approximately same packet delivery ratio when plotted against
offered load.

   C. Node Speed Scenario
Packet Delivery Ratio
Figure 5.21 shows the impact of increasing average speed, with which nodes move, on packet delivery
ratio. Results show that the packet delivery ratio is approximately 100% for all the three protocols and
does not depend upon node speed.

                              100
                              90
  Packet Delivery Ratio (%)




                              80
                              70
                              60
                              50
                              40
                                                                                                         AODV
                              30
                                                                                                         DSR
                              20
                                                                                                         WRP
                              10
                                    5         10             15              20              25
                                                   Node Speed (in m/s)


                                Figure 5.6 Impact of Average Node Speed on packet delivery ratio (FTP traffic)



            6. Conclusion
In this paper, we have simulated the AODV, DSR and WRP routing protocols and evaluated the
performance under CBR and FTP traffics. Performance of each routing protocols evaluated using a
detailed simulation-based analysis. Performance metrics considered are packet delivery ratio.

From the simulation results, for CBR, FTP and TELNET traffics, we concluded the following:

•   For CBR traffic, we have presented 3 different scenarios, varying pause time, offered load (i.e.
number of source destination pairs) and average node speed.
In pause time scenario, performance analysis shows that AODV performs better than DSR and WRP in
terms of packet delivery ratio, throughput and routing message overhead. WRP exhibits the worst
performance in terms of packet delivery ratio, throughput and routing message overhead. But for average
end-to-end delay WRP shows the best performance while DSR shows worst performance in terms of
average end-to-end delay.

•   For FTP traffic, we have presented 3 different scenarios, varying pause time, offered load (i.e.
number of source destination pairs) and average node speed.



                                                                                                                                          399
                                                                                                                  ISSN:2229-6093
     M L Sharma,Noor Fatima Rizvi,Nipun Sharma,Anu Malhan,Swati Sharma, Int. J. Comp. Tech. Appl., Vol 2 (3), 392-400




In pause time scenario, performance analysis shows that DSR performs better than AODV and WRP in
terms of packet delivery ratio and throughput. WRP exhibits the worst performance in terms of packet
delivery ratio, throughput and routing message overhead. But for average end-to-end delay WRP shows
the best performance while DSR shows worst performance in terms of average end-to-end delay AODV
shows the best performance in terms of average end-to-end delay and routing message overhead.
Further Scope
In this section we discuss the further scope of this work.
•       HTTP Traffic
In this work, we have evaluated the performance of routing protocols under CBR, FTP and TELNET
types of traffic. We can evaluate the performance for web server pages i.e. the HTTP traffic.

•       Multimedia Traffic
This work can be extended by evaluating the performance of reactive and proactive protocols under
multimedia traffic. Multimedia traffic is the representative of real time scenario.
.
•      Combining different scenarios
In ad hoc networks, we have analyzed the performance in 3 different scenarios namely, pause time,
offered load and node speed. Further, we can extend this by combining the different scenarios, for
example, varying both pause time and node speed etc. to analyze the performance of each routing
protocol.

•       Considering other scenarios
In ad hoc networks, we have analyzed the performance in 3 different scenarios namely, pause time,
offered load and node speed. Further, we can extend this by evaluating the performance by considering
other scenarios like transmission range, number of nodes etc.

•       Including other routing protocols
In ad hoc networks, we have considered the three routing protocols namely, AODV, DSR (reactive) and
WRP (proactive) routing protocol. We can extend our work by including other reactive or proactive
routing protocols together with some hybrid routing protocols like ZRP.



    7. References
[1] L. Bajaj, M. Takai, R. Ahuja, R. Bagrodia. “GloMoSim: A Scalable Network Simulation
Environment.” Technical Report 990027, University of California, 13, November 1999.

[2] J. Yoon, M. Liu, B. Noble, “Random Waypoint Considered Harmful,” 0-7803-7753-2/03, IEEE
INFOCOM, 2003.
[3] Nipun Sharma “ANALYSIS OF SECURITY REQUIREMENTS IN WIRELESS NETWORKS AND MOBILE
AD-HOC NETWORKS” GESJ: Computer Sciences and Telecommunications 2010 | No.5(28) [2010.11.30]




                                                                                                                        400

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:3
posted:11/18/2011
language:Malay
pages:9