Investigate Performance Analysis Of Routing Protocols

Published Date: 02 Nov 2017

Abstract – Today MANET (Mobile Ad-hoc Network) is a surprisingly impressively changing or growing technology, because of its exclusive nature of distributed nodes and self-motivated topology. The modern and innovative applications based on wireless network are being created in the commercial sectors. The routing protocol follows some standards in wireless MANET’s. These standards control the number of hosts that come to a harmony in such a way that routing packets between all the hosts in the network. The energy efficiency, network stability, security and performance analysis etc are lot of challenges which are facing wireless MANETs. Today MANET’s gain much more attraction due to its use in everywhere in our daily life. Therefore researchers proposed many good routing protocols. In this research article, we investigate performance analysis of some well known wireless routing protocols on the basis of PDR (Packet Delivery Ratio), ROH (Routing Over head), end-to-end delay and throughput.

The wireless networks have rapidly become an essential part of our daily life due to its benefits like accessibility everywhere and any time. The experts and professionals utilize this useful change and increase their work efficiency as well as reducing costs. We give brief introduction about wireless networks and discuss the problems and challenges regarding the performance analysis. We choose three different routing protocols and simulate them in NS-2.34 and investigate different performance related issues. We simulate some well known proactive and reactive routing protocols and finally we have analyzed the results and found the best routing protocol in wireless environments.

Keywords – MANET’s, Routing Protocol, Proactive and Reactive Routing, Simulation of Routing Protocol in NS2, Comparison of Routing Protocol.

Introduction

Today most of the equipments which are equipped with Wireless technology need efficient routing. The routing protocol plays a vital role in wireless environment. MANET is set of hosts that establish relationship between each other without any infrastructure and no centralized management [1]. The hosts are capable of changing their position on regularly basis and interconnect between each other randomly in a MANET. Every node in the network received or forward packets to the nearest node or work as a router. The routing protocol has a main task is to establish correct and efficient route between two hosts in such a way packets can be reached in time.

The creation and maintenance of route should be done under some restrictions like a least amount of overhead and consumption of bandwidth [2].The Wireless MANET has a successful opportunity in the all field of applications. Access and reducing cost are two important issues in the all field of our life. The Bellman-Ford algorithm [3] is used in DSDV proactive routing protocol. The combination of route discovery and route maintenance is the procedure of DSR [4]. The whole process repeat, during searching the entire route path the destination host sends route reply message to source host for building successful route finding method [5].

Mobile Ad-hoc Routing Protocols

Figure 1

MANET routing protocol have three types.

Proactive MANET Protocols

Reactive MANET Protocols

Hybrid MANET Protocols

1. Proactive MANET Protocols

All the hosts constantly look for routing information with in the network, in order that the route is already identified, if a route is required. When a host needs to communicate to another host, the path is recognized so that latency is small and the maintenance cost of all topology information is maximum if large number of host movements [6].

2. Reactive MANET Protocols

At any time if a host is need of a path from source to destination at that time an inquiry reply acknowledgement procedure start, therefore latency is high and no need of unwanted control message [7, 8].

3. Hybrid MANET Protocols

The mixture of proactive and reactive method is called hybrid protocol. Today researchers are proposed many hybrid protocols like ZRP, CEDAR and ZHLS etc [9]. Researchers simulate and investigate many routing protocols on the basis of performance in different simulation environment like NS2, OMNET++, OPENET and GloMoSim etc. In order to simulate and evaluate the routing protocols, we selected DSDV from proactive, DSR and AODV from the reactive.

II. Related work

Different routing protocols have been produced with the help of simulation software like NS2, QualNet and GLOMOSIM. Birdar et al. presented in his paper a comparison of performance analysis for AODV and DSR in NS2 [10] using speed as a variable parameter. The G. Jayakumar et al. also presented in his paper a comparison of performance analysis for AODV and DSR in NS2 [11] using number of sources, Speed and pause time as a variable parameter. Yogesh et al. in his paper presented a performance analysis for AODV and DSR in GLOMOSIM [12] using variable parameter as number of node, speed and pause time. Shaily et al. compare AODV, DSR and ZRP using variable parameter as pause time in QualNet [13]. Vijayalaskhmi et al. in his paper analyze performance of DSDV and AODV using variable parameter as pause time in NS2 [14].

III. Simulations

The Reference Point Group Mobility model (RPGM)

has been used with node speed between 0.5 to 5.0 m/s, simulation time is 300 seconds, transport protocols is UDP and traffic generator source is CBR. The node density and simulation area varies from 20 to 80 nodes and 500m x 500m. The two-ray ground is used as propagation model. The other limitation used in simulation is written in table 1.

Simulation Parameters

Transmission range

500m x 500 m

Antenna

Omni directional

MAC Type

IEEE 802.11

Node speed

0.5m/s to 5.0 m/s

Simulation Time

300 sec

Packet rate

8 packet/sec

Traffic Type

CBR

Data payload

512 bytes/ packet

Interface Queue Type

Drop Tail/Priori Queue

Node Pause Time

0

Mobility Model

RPGM

Interface Queue Length

50

No. of Nodes

20, 40, 60, 80

Table 1

Some parameters in table 1 are constant and some are variable. During simulation these parameters are varying to test and verify the outcome (number of node and node speed).

Performance Matrix

To measure the performance of protocols is evaluated on some parameters which are End to End Delay, Packet Delivery Ratio, Routing Overhead, and Throughput.

1. PDR/ PDF (Packet Delivery Ratio/ Function)

A ratio between the numbers of sent packet from the source node and numbers of received packet to the receiver node is called PDR / PDF.

Where P is packet delivery ratio / function, Pkt_S is packet send and Pkt_R is packet receive.

2. ROH (Routing Overhead)

The number of nodes as counting the hop-wise communication of packet which is considered as single communication is called ROH.

Where R is Routing Overhead and No. of RTR is

3. Throughput

Throughput is defined as total no. of data packet that is effectively received at the receiving node dividing by total no. of data packet sending through the network.

Where T is throughput, Pkt. Rec is packet received and Pkt. Snd is packet send.

4. Average E2E Delay (Average End to End Delay)

This performance metric include all possible delays during communication. During network communication there are many factors causing delay, such as, latency, buffering during routes discovery, queuing delay, and retransmission delay. Better performance means minimum delay.

Ri = time at which data packet i was received

Si = time at which data packet i was sent

K = Total number of received data packets.

IV. Results

The performance analysis of DSR, DSDV and AODV protocols are evaluated in term of PDR, ROH, Throughput and Average end-to-end delay. The node density varies from 20 nodes to 80 nodes.

PDR: According to the figure 1 it is proved that packet delivery ratio is maximum in AODV protocol. PDR is minimum in DSR from node 20 to 60 and minimum from node 60 to 80. PDR is medium from node 20 to 60 in DSDV and minimum from node 60 to 80.

Figure 2

Throughput: In figure 2 it is evident that throughput of AODV from node 20 to 80 is maximum but in DSR from node 20 to 40 is maximum and 40 to 60 is medium and 60 to 80 is minimum On the other hand throughput of DSDV from node 20 to 60 is minimum and from node 70 to 80 is medium.

Figure 3

Average End-to-End Delay: According to the figure 3 it is observed that Average End-to-End Delay is minimum in AODV protocol, medium in DSDV and maximum in DSR.

Figure 4

ROH: In figure 4 it is obvious that ROH is maximum in AODV but in DSDV from node 20 to 40 and 70 to 80 is medium and from node 40 to 60 is minimum. On the other hand ROH in DSR is minimum from node 20 to 40 and 60 to 80 and medium from node 40 to 60.

Figure 5

It is interesting to see that performance of DSR in ROH is best among all the three routing protocols. DSDV gives comparatively medium performance in all simulation results. Because DSDV is a proactive routing protocol, performs table update operations from time to time as compared to the reactive protocols. It is evident from the above figures the performance is not dependent on node density. AODV is the next better protocol and at high node density it is even better than DSDV and DSR. Overall AODV performs better than other protocols in the simulated scenario. The performance analysis of AODV is reasonable in all the four metrics, which makes it clearly the less expensive protocol.

V. Conclusion

This research paper gives a summary of Wireless Ad-hoc network and discusses how performance is one of the most essential constraints for these types of networks. A comprehensive study of the performance analysis strategies and performance analysis metrics is provided. According to this study it is seen that focusing on four performance analysis metrics for routing in order to achieve better performance. We evaluate and simulate three routing protocols to investigate the performance analysis of routing protocols and find that AODV gives best performance in Average end-to-end delay, throughput and PDF but in ROH DSDV is best. The objective of thiswork is to develop an efficient performance routing algorithm in a way which allows researchers to choose the most appropriate routing algorithm.

Acknowledgement

We are grateful to Department of Computing and technology, IQRA University Islamabad Campus (Pakistan) for providing necessary facilities and peaceful environment to complete this research work. We also presents humble thanks to our research supervisor Dr. Imran Shafi for providing outstanding supervision and support to complete this document in its present form. We would like to thank Mr. Amjad Arfin (Lab Administrator) and family members for financial and moral supports. Finally, we are also grateful to our senior Ph. D. research scholar Mr. Qamar and colleagues who have provided technical and academic guidance during this research work.