comparative analysis and performance evaluation of routing … · 2016-04-11 · routing (t-dsr),...

International Journal of Advanced Computational Engineering and Networking, ISSN: 2320-2106, Volume-2, Issue-9, Sept.-2014

Comparative Analysis And Performance Evaluation Of Routing Protocols In Manet

50

COMPARATIVE ANALYSIS AND PERFORMANCE EVALUATION OF ROUTING PROTOCOLS IN MANET

1AMIT M. KAVATHEKAR, 2SANDIP A. THORAT

1M.Tech Student, Department of CSE, Rajarambapu Institute of Technology

2Associate Prof, Department of IT, Rajarambapu Institute of Technology, Islampur (Sangli), MS, India E-mail: [email protected], [email protected]

Abstract- A Mobile Ad hoc NETwork (MANET) is network of independent mobile nodes which forms at run time. It is collection of wireless mobile nodes organized to create a temporary connection between them. Over the last few years, many routing protocols for MANETs have been proposed and enhanced to route data packets between two nodes in network. It is not clear, however, how different protocols behave in different environments. Any one protocol may be good to use in one network topology and mobility scenario, but to worst to use in another. This paper focuses on three types of routing protocols named Dynamic Source Routing (DSR), the traditional protocol; ARIADNE, the secured protocol and Trust based Dynamic Source Routing (T-DSR), trusted protocol. The analysis takes DSR as a basic protocol and does its performance comparison with rest of other two protocols with respect to packet delivery ratio, control overhead and average end to end delay. I. INTRODUCTION A Mobile Ad hoc Network is self-organized network of independent mobile nodes without requiring any kind of fixed infrastructure. Because of some of its features such as openness, mobility, dynamic topology, rapid deployment and protocol weaknesses, MANETs are prone to be unstable and vulnerable. Consequently, their security requirements are very urgent and it is not so easy to design and implement security solutions for MANETs than for wired networks. Traffic analysis attack is one of the unsolved security attacks against such networks. Although the secure routing for MANETs has been effectively studied, traffic analysis attacks are still not well addressed. Though protocols focus on security of route maintenance, they cannot prevent traffic analysis attack. A number of secure routing protocols have been proposed, which make use of cryptographic algorithms to secure the routes. However, they need number of prerequisites during both network establishment and operation phases. In contrast, trust- based routing protocols locate trusted routes in the network by observing the behaviour of participating nodes. There are kind of trust based protocols which are totally based upon trust mechanisms and never rely on cryptographic schemes. Use of trust concepts plays vital role in MANET security. In the rest of the paper, we first discuss some relevant previous work in Section II. In Section III, we briefly reported the overview of routing protocols considered for analysis with their design and methodology used for implementation. In Section IV, we describe the simulation environment and simulation results for random scenario are given and observations are noted down based on the results obtained, with concluding remarks in Section V.

II. RELATED WORK This section covers review of performance analysis of different routing protocols in MANET. Karan Singh et al. presents a survey of secure ad hoc routing protocols for wireless networks. They briefly present the most popular protocols that follow the table-oriented and the on-demand approaches. The comparison between the proposed solutions and parameters of ad hoc network shows the performance according to secure protocols. Asad Amir Pirzada et al. presented comparative performance analysis of three trust-based routing protocols in an attacked ad hoc network with malicious nodes. R.G. Rajkumar et al. presented routing protocols DSR and AODV. Performance of these two protocols has been compared for large number of nodes. From their results it is evident that AODV protocol consumes lesser energy and in the presence of high routing overhead, AODV outperforms DSR by having higher throughput and less jitter. Poonam, K. Garg et al. presented a novel method based on trust value which is used to find a secure path from source to destination free from malicious nodes in DSR routing. In this method, each node broadcasts a RREQ packet if it is received from different neighbors. Therefore, at the destination we have multiple paths with different nodes, which further lead to the discovery of the highly secure path, avoiding compromising nodes. Yih-Chun Hu et al. presents attacks related to ad hoc networks, and they present the design and performance evaluation of a new secure routing protocol, called Ariadne. Ariadne prevents compromised nodes from forming the routes consisting of compromised nodes, and also prevents from Denial-of-Service attacks. Ariadne provides security against compromised node and relies only on



51

symmetric cryptography. Ariadne protocol is being discovered routes between nodes only as needed. III. PROTOCOL DESCRIPTION This section gives short description of three routing proto-cols included in this paper. A. DSR : Dynamic Source Routing protocol Dynamic Source Routing (DSR) is an on demand reactive routing protocol. In DSR, the routes are established only on-demand. Table-driven approach is not considered in reactive routing protocols. It doesn’t use beacon packets to inform its neighbors about its existence. DSR is a source routing protocol, which means the entire route is known before a packet transmission has begun. DSR stores discovered routes in a route cache. It uses two mechanisms i.e. Route discovery and Route maintenance. In route discovery, it sends route request i.e. RREQ packet to all nodes in the network, where RREQ packet is a broadcast packet. On receiving the RREQ packet at destination it responds to the source node with a route reply i.e. RREP packet in the reverse path of RREQ packet. B. ARIADNE: A Secure on demand routing protocol for adhoc networks. ARIADNE is an ad hoc on-demand secure routing protocol, in which it can work against compromised nodes and based only on symmetric cryptography. It applies and extends DSR, in the sense that only the source and the destination nodes can verify the properness of the established route. The intermediate nodes which participating the route discovery phase can’t get any confirmed route from it. Ariadne can authenticate routing messages using one of three schemes i.e. shared secret keys between all pairs of nodes, shared secret keys between communicating nodes combined with broadcast authentication, or digital signatures. We have been focused here the use of Ariadne with TESLA (Time-Efficient Stream Loss-Tolerant Authentication for ad hoc networks), an efficient broadcast authentication scheme which needs loose time synchronization. In this paper, we shortly describe Ariadne protocol using the TESLA broadcast authentication mechanism for authenticating routing messages, since TESLA is efficient and adds only a single message authentication code (MAC) to a message for broadcast authentication. In Ariadne Route Discovery using TESLA broadcast authentication, every source-destination pair of nodes S and D share the MAC keys KSD and KDS. Route Discovery contains two stages: the initiator floods the network with a ROUTE REQUEST, and the target returns a ROUTE REPLY. To secure the ROUTE REQUEST packet, Ariadne provides the following properties: the target node can authenticate the initiator using a MAC with a key shared between the

initiator and the target; the initiator can authenticate each entry of the path in the ROUTE REPLY (each intermediate node appends a MAC with its TESLA key); and no intermediate node can remove a previous node in the node list in the REQUEST or REPLY (a one-way function prevents a compromised node from removing a node from the node list). C. T-DSR: Trust based multipath routing protocol for ad hoc networks. 1) T-DSR protocol Architecture: T-DSR is Trust based multipath routing protocol for ad hoc networks. It is an extension of the DSR protocol. The protocol discovers multiple paths between two communicating nodes. This is essential for an ad hoc network to be able to tolerate attack-induced path failures. Fig.1 shows T-DSR protocol architecture in which monitor is responsible for registering packets which are sent by a node and supposed to be forwarded by the next hop. Rating System maintains the ratings of other nodes based on the packet forwarding ratio published by other nodes. Path maintenance is responsible for maintaining cached routes and selecting a safe route which contains no misbehaved nodes. Trust design component implements methods to assign trust values to nodes when they are first encountered. Nodes will be encountered from ROUTE REPLYs or when outes are snooped. When the first routes are discovered, all nodes on the route will be unknown and therefore default trust value for unknown nodes need to be determined.

Fig. 1. T-DSR protocol Architecture.

2) Trust updating: The trust updating component implements the functions for updating trust. The function designed here aims to function in domains with several malicious nodes. Ideally the behavior of



52

the function should depend on the expected trust dynamics in a given situation. Since the trust values are used to base decision on which route to choose, only one function is used.

IV. SIMULATION AND ANALYSIS A. Simulation environment In this work, we have carried out extensive simulations using network simulator NS-2.34 and the performance analysis were conducted by AWK scripts for different performance parameters. For the evaluation of different parameters the average value of performance results of five random scenario patterns were considered. All simulations were performed on Ubuntu 10.04. We used Random waypoint mobility model to determine the node mobility and related attributes. Table I provides all the simulation parameters.

TABLE I SIMULATION PARAMETERS

B. Simulation results and analysis In this section, we display results of DSR, ARIADNE and T-DSR protocols by conducting different simulation tests. Malicious nodes are also present in the network. The graphs are plotted using GNU Plot software which shows comparison. We used following performance metrics to test the performance of DSR, ARIADNE and T-DSR protocols. • Packet delivery ratio: The ratio of the data packets delivered to the destination to those generated by CBR or FTP sources [19]. • Average End to end delay: It gives mean time in seconds taken by data packets to reach their respective

destinations. It includes all possible delays in the network caused by route discovery latency, retransmission by intermediate nodes, processing delay, queuing delay and propagation delay. To average the end-to-end delay we add every delay for each successful data packet delivery and divide that sum by no. of successfully received data packets [4]. • Control overhead: It is the ratio between total no. of control packets generated to the total no. of data packets received during simulation time [4]. We evaluate the performance of DSR, ARIADNE and T-DSR by using following simulation tests: 1) Varying number of nodes in the network The number of nodes varies from 50 to 250 nodes (i.e. 50,100,150,200,250), with node mobility speed is 10 m/s, pause time is 2 sec, transmission range is 250m and 2% of malicious nodes. 2% of malicious nodes in the sense 2% of total no. of nodes (e.g. if no. of nodes have been taken 100 then 2% of 100 is 2 malicious nodes). Therefore, if number of nodes in the network going to increase then no. of malicious nodes in the network alsogoes on increasing. Figure 2 shows Packet delivery ratio comparison between DSR, ARIADNE and T-DSR under varying no. of nodes in the network. We can see that the packet delivery ratio of T-DSR protocol is apparently higher than that of ARIADNE and DSR. This attributes to effectiveness of malicious node detection algorithm used in T-DSR when malicious node increases. The backup paths in DSR protocol is intersecting and the broken place is likely to be public node of two paths, thus this lead to larger failure probability of the backup path restoration and further cause higher loss probability of data packet. In ARIADNE, due to key generation process the packet delivery ratio will low. But T-DSR protocol can quickly find alternate paths to arrive the same destination node, so the probability of packet discarded is small. At the time of the malicious node enters, T-DSR protocol can find alternate paths of non-existent public node. So the packet delivery ratio will be high in T-DSR protocol as compared to DSR protocol and ARIADNE protocol.

Fig. 2. Packet delivery ratio versus no. of nodes in network.



53

Figure 3 shows average end to end delay comparison between DSR, ARIADNE and T-DSR under varying no. of nodes in the network. We can see that while no. of nodes increases the average end to end delay of three protocols all increased, but the average end to end delay of T-DSR protocol is lower than DSR and ARIANDNE protocols. In ARIADNE we have used key generation process and authentication process. So the overhead will be high in ARIADNE protocol. So, average end to end delay will be very high regarding ARIADNE protocol. Figure 4 shows control overhead comparison between DSR, ARIADNE and T-DSR under varying no. of nodes in the network. For ARIADNE protocol, while number of nodes increases the route request message RREQ takes up 90% of the total number of control messages. When node increases re-route discovery of source node, all will lead to transmit a lot of RREQ message in network. Therefore, the overhead will significantly increase with the network topology drastic change. But in the course of a route discovery, T-DSR

protocol can get more multipath routing. In the case of malicious node enters inside network, source node can quickly find an alternate route as backup path and continue to transfer data, thus there is no need to perform route discovery, as a result it significantly reduce the number of control message.

2) Varying number of malicious nodes in the network We are going to vary the no. of malicious nodes in the network. The malicious nodes varies from 2% to 10% (i.e. 2%, 4%, 6%, 8%, 10%).We have taken the malicious nodes in percentage in the sense e.g. 2% of

malicious nodes is equal to 2% of total no. of nodes present in the network. We are going to vary malicious nodes in the network with no. of nodes are 50, transmission range is 250m, pause time is 2 sec and node mobility speed is 10m/s. As we have taken total 50 nodes in the network, malicious nodes varies from 2% to 10% that means 1 to 5 malicious nodes (1, 2, 3, 4, 5 malicious nodes). Figure 5 shows packet delivery ratio comparison between DSR, ARIADNE and T-DSR under varying malicious nodes in the network. We can see that the packet delivery ratio of T-DSR protocol is apparently higher than that of ARIADNE and DSR protocol. The backup paths in DSR protocol is intersecting and the broken place is likely to be public node of two paths, thus this lead to larger failure probability of the backup path restoration and further cause higher loss probability of data packet. And in ARIADNE protocol, due to key generation process the packet delivery ratio will low. But T-DSR protocol can quickly finds alternate paths to arrive the same destination node, so the probability of packet discarded is small. At the time of the malicious node enters, T-DSR protocol can find alternate paths of non-existent public node.so the packet delivery ratio will be high.

Figure 6 shows average end to end delay comparison between DSR, ARIADNE and T-DSR under varying no. of malicious nodes in the network. We can see that while malicious node increases the average end to end delay of three protocols all increased, but the average end to end delay of T-DSR protocol is lower and smoother than that of ARIANDNE and DSR protocol. In particular, when the malicious node increases the differences between T- DSR, ARIADNE and DSR protocol is growing wider. When malicious nodes are maximum, average end to end delay of T-DSR protocol is low to nearly 20% than that of ARIADNE protocol and lower compared with the DSR protocol. This is mainly because T-DSR protocol can use alternate routing if it identified the malicious node so then reduced the delay of source node rerouting. Moreover T-DSR protocol adopts minimum hops path in selecting path to be great influence on average end-to- end delay.



54

Figure 7 shows Control overhead comparison between DSR, ARIADNE and T-DSR under varying no. of malicious nodes in the network. For ARIADNE protocol, when malicious node enters, whether local restoration or re-route discovery of source node, all will lead to transmit a lot of RREQ message in network. Therefore, the control overhead will significantly increase with the network topology drastic change. But in the course of a route discovery, T-DSR protocol can get more multipath routing. In the case of malicious node enters inside network, source node can quickly find an alternate route as backup path continue to transfer data, thus there is no need to perform route discovery, as a result it significantly reduce the number of control message. So the control overhead in T-DSR protocol is lower than that of ARIADNE protocol. 3) Varying mobility speed of nodes in the network The mobility speed varies from 10m/s to 50m/s (i.e.10, 20, 30, 40, 50m/s), with pause time is 2 sec, transmission radius of node is 250m, number of nodes are 50 and 2%

of malicious nodes present in the network (i.e.2% of 50 is equal to 1 malicious node is present). Here no. of malicious nodes remains constant (i.e.1 malicious node) because no. of nodes (i.e.50 nodes) being remain constant each time in network. We compare the performance of DSR, ARIADNE and T-DSR under varying mobility speed of nodes in the network. Figure 8 shows packet delivery ratio comparison between DSR, ARIADNE and T-DSR under varying mobility speed of nodes in the network. We can see that the packet delivery ratio of T-DSR protocol is apparently higher than that of ARIADNE and DSR protocol. When network topology change speed up, the success rate of ARIADNE will reduced and lead to

cache data packet loss. The backup paths in DSR protocol is intersecting and the broken place is likely to be public node of two paths, thus this lead to larger failure probability of the backup path restoration and further cause higher loss probability of data packet. But T-DSR protocol can quickly find alternate paths to arrive the same destination node, so the probability of packet discarded is small. At the time of the malicious nodes enters, T-DSR protocol can find alternate path, therefore, the chance of path break due to malicious node is smaller, so the packet delivery fraction increased. Figure 9 shows average end to end delay comparison between DSR, ARIADNE and T-DSR under varying mobility speed of nodes in the network. We can see that with the acceleration of the network topology changing the average end to end delay of three protocols all increased, but the average end to end delay of T-DSR protocol is lower and smoother than that of DSR and

ARIADNE protocols. In particular, with the network topology change intensified the differences between T- DSR, ARIADNE and DSR protocol is growing wider. When a node maximum movement speed is to 15 m/s, the delay of T-DSR protocol is low to nearly 70% than that of ARIADNE protocol and lower compared with the DSR protocol. This is mainly because T DSR protocol can use alternate routing in malicious nodes enter so then reduced the delay of source node rerouting.

Figure 10 shows control overhead comparison between DSR, ARIADNE and T-DSR under varying mobility speed of nodes in the network. In case of ARIADNE protocol, the control overhead drastically increases as the mobility speed increases. The control



55

overhead also increases as the node mobility increases. Comparing to three protocols, T-DSR protocol having low control overhead than other two protocols.

CONCLUSION AND FUTURE WORK In this paper, we have done comparative performance analysis of DSR, ARIADNE and T-DSR protocols in presence of malicious nodes under different kind of simulation tests. We have conducted these tests by varying different network parameters (varying nodes, malicious nodes, node mobility speed and transmission radius) mentioned in previous section. We have measured the performance by using different performance metrics which are end to end delay, control overhead and packet delivery ratio. After applying different tests we have observed that T-DSR protocol has better performance than DSR and ARIADNE protocols. The performance of T- DSR protocol is apparently higher than that of ARIADNE and DSR protocols. This is mainly because T-DSR protocol can use alternate routing if it identified the malicious node. Future work can expand the trust model used in T-DSR protocol where trust value evaluation of node will be more accurate. Instead of packet forwarding ratio we can use other more effective performance parameter for rating the nodes in the network. Trust updating mechanism used in T-DSR protocol can be changing for more accurate and efficient neighbor node selection for forwarding the packet to the destination. We can compare this modified trust based protocol with any other protocols under different performance metrics such as average hop count, route acquisition time, packet loss, path optimality etc. REFERENCES

[1] Xiaoqi Li; Lyu, M.R.; Jiangchuan Liu; ”A trust model based routing protocol for secure ad hoc networks” Aerospace Conference, 2004 Proceedings. 2004 IEEE, vol.2, no., pp. 1286- 1295 Vol.2, 6-13 March 2004.

[2] Elahe Sheklabadi; Mehdi Berenjkoub;”An Anonymous Secure Routing Protocol for Mobile Ad Hoc Networks” International symposium on computer networks and distributed systems(CNDS), February 23-24, 2011.

[3] Karan Singh; R. S. Yadav; Ranvijay ;”a review paper on ad hoc network security” International Journal of Computer Science and Security, Volume (1): Issue (1).

[4] Asad Amir Pirzada; Chris McDonald; Amitava Datta; ”Performance Comparison of Trust-Based Reactive Routing

Protocols” IEEE trans- actions on mobile computing, vol. 5, no. 6, june 2006.

[5] Pirzada A.A.; McDonald C. ”Trusted route discovery with TORA protocol” Communication Networks and Services Research, 2004 Pro- ceedings. Second Annual Conference on, vol., no., pp. 121- 130, 19-21 May 2004.

[6] Rashmi Rohankar; Rinkoo Bhatia; Vineet Shrivastava; Deepak Kumar Sharma; ”Performance Analysis of Various Routing Protocols (Proactive and Reactive) for Random Mobility Models of Ad hoc Networks” 1st Int’l Conf. on Recent Advances in Information Technology - RAIT-2012.

[7] Rajkumar G.; Kasiram R.; Parthiban D ”Optimized QoS metrics and performance comparison of DSR and AODV routing protocols” Advances in Engineering, Science and Management (ICAESM), 2012 International Conference on , vol., no., pp.251-255, 30-31 March 2012.

[8] Poonam; Garg K.; Misra M.”Trust Based Multi Path DSR Protocol” Availability, Reliability, and Security, 2010. ARES ’10 International Conference on , vol., no., pp.204-209, 15-18 Feb. 2010.

[9] Mohamad Y.Alsaadi Yi Qian ”Performance Study of a Secure Routing Protocol in Wireless Mobile Ad Hoc Networks”1-4244-0523-8/07/20.002007IEEE.

[10] Jared Cordasco, Susanne Wetzel, ”Cryptographic Versus Trust-based Methods for MANET Routing Security” Electronic Notes in Theoretical Computer Science, Volume 197, Issue 2, 22 February 2008, Pages 131-140.

[11] Imani M.; Taheri M.; Rajabi M.E.; Naderi M.;”A secure method on a routing protocol for ad hoc networks” Educational and Network Technology (ICENT), 2010 International Conference on vol., no., pp.482-486, 25-27 June 2010

[12] Y.-C. Hu, A. Perrig, and D. B. Johnson,”Ariadne: a secure on-demand routing protocol for ad hoc networks,” Wirel. Netw., vol. 11, no. 1-2, pp. 21-38, 2005.

[13] Abusalah L.; Khokhar A.; Guizani M.; ”A survey of secure mobile Ad Hoc routing protocols”, Communications Surveys Tutorials, IEEE, vol.10, no.4, pp.78-93, Fourth Quarter 2008

[14] Jalel Ben Othman, Lynda Mokdad, ”Enhancing data security in ad hoc networks based on multipath routing”, Elsevier Inc.,2009, doi:10.1016/j.jpdc.2009.02.010.

[15] Praveen Joshi, ”Security issues in routing protocols in MANETs at network layer ”, Elsevier2010, doi:10.1016/j.procs.2010.12.156.

[16] Todd R Andel, Air force institute of technology Alec Yasinsac, Florida state university,”Serveying security analysis techniques in MANET routing protocols”, IEEE Communications Surveys Tutorials 4th Quarter 2007.

[17] P.Srinivasan1, K. Kamalakkannan, ”Signal Strength and Energy Aware Reliable Route Discovery In Manet”, International Journal Of Communication Network Security ISSN: 2231 - 1882, Volume-1, Issue-4, 2012.

[18] Amar M. Hassain; Mohamed I. Youssef; Mohamed M. Zahra, ”Evaluation of Ad Hoc Routing Protocols in Real Simulation Environments”, IEEE 2006.

[19] Putta, C.S.R.; Prasad, K.B.; Ravilla, D.; Nath, R.S.M.; Chandra, M.L.R., ”Performance of ad hoc network routing protocols in IEEE 802.11,” Computer and Communication Technology (ICCCT), 2010 International Conference on , vol., no., pp.371,376, 17-19 Sept. 2010

[20] Govindan, K.; Mohapatra, P., ”Trust Computations and Trust Dynamics in Mobile Ad hoc Networks: A Survey,” Communications Surveys Tutorials, IEEE , vol.14, no.2, pp.279,298, Second Quarter 2012.

comparative analysis and performance evaluation of routing … · 2016-04-11 · routing (t-dsr),...

Documents