Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide


  1. 1. Pradish Dadhania, Sachin Patel / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 3, Issue 1, January -February 2013, pp.1487-1491Performance Evaluation of Routing Protocol like AODV and DSR under Black Hole Attacks Pradish Dadhania*, Sachin Patel** *(Department of Information Technology, RGPV University, Indore (INDIA) ** (Department of Information Technology, RGPV University, Indore (INDIA)ABSTRACTThe Wireless mobile ad hoc network (MANET) is used in many areas. Such as in disaster hit areas,is a self-configuring network which is composed data collection in some region, in rescue missions,of several movable mobile nodes. These mobile virtual classes and conferences [8]. This conceptnodes communicate with each other without any with ad hoc network makes the full name of mobileinfrastructure. As wireless ad-hoc networks lack ad hoc network (MANET). By growing thean infrastructure, they are exposed to a lot of network, combined with the node mobility theattacks. One of these attacks is the Black Hole challenges of self configuration of the networkattack. In Black Hole attack, a malicious node become more evident.falsely advertises shortest path to the destinationnode and absorbs all data packets in it. In this Security in MANET is a very critical and importantway, all packets in the network are dropped. In issue and many techniques were defined for thethis paper, performance of AODV and DSR are security of MANET. Intrusion detection techniqueevaluated in presence of black hole attack is investigated in [8].Mobile nodes in the network(malicious node) and without black hole attack waste much energy by joining in and out withwith CBR traffic under different scalable connection to wireless network. This connectionnetwork mobility. and reconnection create energy limitation in theIn this paper, via simulation, we evaluate effect wireless network. The main purpose of developingand compare it with standard protocol in terms the ad hoc routing protocols is to cope with theof throughput, Packet delivery ratio and End to dynamic nature of MANET. The routing protocolsEnd Delay. We have conducted extensive efficiency can be determined by the battery powerexperiments using the network simulator-2 to consumption. Energy is consumed duringvalidate our research. participation of a node in a network and also in routing of traffic.Keywords - MANET, ADOV, DSR, Black hole, There are mainly three types of routing protocolI. INTRODUCTION and they are proactive, reactive and hybrid routingThe mentioned before an ad hoc network is a protocol. Proactive routing protocol is table drivenwireless network, which do not have a centralized where as reactive routing protocol is on demandand fixed infrastructure. MANET is referred to as a routing protocol. AODV and DSR both are onwireless ad hoc network in which nodes are free to demand protocol. But the difference between bothmove arbitrarily and mobile nodes can transmit and of these is, DSR a route cache is maintained andreceive the traffic. Also mobile nodes can act like due to this over head of memory increases. But inrouters by forwarding the neighbors traffic to the case of AODV, it is a source initiate routingdestination node as the routers are multi hop protocol. I n this, routing table is maintained bydevices [8]. MANET does not need base stations of every mobile node and this routing table consist ofwired infrastructure. The mobile nodes in wireless next node information for a route to the destinationnetwork range can communicate with each other node. The intermediate nodes between the sourcebecause it is a self organized network. The mobile and destination are responsible for finding a freshnodes form a network automatically without a fixed path to the destination in route discovery process ofinfrastructure and central management [8]. The AODV protocol. Malicious node immediatelymobile nodes have transmitters and receivers with responses to such route discovery process givingsmart antennas, which enable the mobile nodes to false information of having a fresh enough path tocommunicate with each others. destination. Source node assumes that it is sending data packets through a true path but actually itThe topology of the network changes every time by sending the data packets to malicious node. Apartgetting in and out of the mobile nodes in the from the malicious node, black hole attack cannetwork. In the beginning MANET was designed occur due to damaged node, unintentionallyfor military use but now the MANET dropping of data packets. 1487 | P a g e
  2. 2. Pradish Dadhania, Sachin Patel / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 3, Issue 1, January -February 2013, pp.1487-1491 source route. It may send that reply along theII. OVERVIEW OF AODV AND DSR source route in reverse order or issue a ROUTE ROUTING PROTOCOLS REQUEST including the route to get back to the source, if the former is not possible due to2.1 The Ad hoc on-Demand Distance Vector asymmetric links. ROUTE REPLY messages can PROTOCOL be triggered by ROUTE REQUEST messages orAODV is on demand routing, means it start its are gratuitous. After receiving one or severalrouting process only when any node in the network routes, the source selects the best (by default thedesire to transmit the data packets. In AODV, next shortest), stores it, and sends messages along thathop information is started by each node in it a path. The better the route metrics (number of hops,routing table. When a source node cannot reach to delay, bandwidth, or other criteria) and the soonerthe destination node directly, then the source node the REPLY arrives at the source, the higher thewill immediately initiate a route discovery process. preference given to the route and the longer it willAODV uses several control packets like Route stay in the cache. When a ROUTE REPLY arrivesRequest (RREQ), Route Reply (RREP) and Route very quickly after a ROUTE REQUEST has beenError Process (RERR). RREQ message is sent out this is an indication of a short path, sincebroadcasted, RREP message is unicasted back to the nodes are required to wait for a timesource of RREQ, and RERR message is used to corresponding to the length of the route they cannotify the loss of link to other node. Route advertise, before sending it. This is done in order todiscovery is initiated by broad casting a RREQ to avoid a storm of replies. In case of a link failure,its neighbor and this RREQ is rebroadcasted to the node that cannot forward the packet to the nexttheir neighbor until it reaches to the destination node sends an error message towards the source.node. When destination node receives the RREQ, it Routes that contain a failed link can be `salvagedsends the RREP message to the sender node Routes by taking an alternate partial route that does notare maintained in the source node as long as they contain the bad link.are needed. Routing table are maintained by everynode and have fields like destination, number of 3 BLACK HOLE ATTACKS ONhops, next hops, destination sequence number, life MANETtime, active neighbor. To find the freshness ofroute towards destination, sequence number is MANETs face various securities threats i.e. attackused. Attacks on AODV can be performed easily as that are passed out against them to interrupt theAODV does not have any centrally administered normal performance of the networks. Black holesecure routers. Attackers from outside or inside can attack is one of the security threat in which theeasily exploit the network. AODV supports shared traffic is redirect to such a node that actually doeswireless medium and dynamic topology. It is not exist in the network. In these attacks, black holecapable of both unicast and multicast routing. It attack is that kind of attack which occurs in Mobileavoids count to infinity problem of other Distance Ad-Hoc networks (MANET). In black hole attack,–vector protocol. It is flat routing protocol and does a malicious node uses its routing protocol in ordernot need any central administrative system to to endorse itself for having the shortest path to thehandle the routing process. It does not require any destination node or to the packet it wants topermanent link between the nodes to transfer data. interrupt. This destructive node advertises itsFor transferring the data temporary link would availability of new routes irrespective of checkingsuffice for time being. AODV needs less protection its routing table. In this way attacker node willof control message. It is enough to protect RREQ always have the availability in replying to the routeand RREP message in order to provide the security request and thus intercept the data packet and retainto the protocol. it [6]. In protocol based on flooding, the malicious2.2 Dynamic Source Routing protocol node reply will be received by the requesting node before the response of reply from actual node;Dynamic Source Routing is a protocol developed hence a malicious and forged route is created.for routing in mobile ad-hoc networks and was When this route is establish, now it is up to theproposed for MANET by Broch, Johnson, and node whether to drop all the packets or promote itMaltz [2]. In a nutshell, it works as follows: Nodes to the unknown address [7]. The black hole attacksend out a ROUTE REQUEST message, all nodes has two properties. First, the node exploits thethat receive this message put themselves into the mobile ad-hoc routing protocol, such as AODV, tosource route and forward it to their neighbors, promote itself as having a valid route to a targetunless they have received the same request before. node, even though the route is false, with the aimIf a receiving node is the destination, or has a route of intercepting packets. Second, the attackerto the destination, it does not forward the request, consumes the intercepted packets without anybut sends a REPLY message containing the full forwarding. However, the attacker runs the risk that neighboring nodes will check and represent the 1488 | P a g e
  3. 3. Pradish Dadhania, Sachin Patel / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 3, Issue 1, January -February 2013, pp.1487-1491ongoing attacks. There is a more delicate form of size, the overall performance of the protocols canthese attacks when an attacker selectively forward be compared in terms of three parameters:packets. An attacker suppress or modifies packetsoriginating from some nodes, while leaving the 5.1.1 Throughputdata from the other nodes unchanged, which limits Throughput or network throughput is the averagethe suspicion of its wrongdoing. rate of successful message delivery over a communication channel. The throughput is usually measured in bits per second (bit/s or bps), and sometimes in data packets per second or data packets per time slot. From graph we can analyze as number of node increase in network throughput gets better. Fig: 1 - BLACK HOLE ATTACKS4 SIMULATION SET UPThe simulation is implemented In NetworkSimulator 2 [8], a simulator for mobile Adhocnetworks. The simulation parameters are providedin Table 3. We implement the random waypointmovement model for the simulation, in which anode starts at a random position, waits for thepause time, and then moves to another randomposition with a velocity chosen between 0 m/s to Fig: 2 throughput Of DSR(Sesr.1) andthe maximum simulation speed. A packet size of AODV(ser.2)512 bytes and a transmission rate of 4packets/s,congestion of the network are not likely 5.1.2 Average end-to-end delay of data packetsto occur. There are possible delays caused by bufferingParameter Value during route discovery latency, queuing at theExamined Protocol AODV DSR interface queue, retransmission delays at the MAC,Application traffic CBR and propagation and transfer times. Once the timeTransmission range 250 m difference between every CBR packet sent andPacket size 512 bytes received was recorded, dividing the total timeTransmission rate 4 packets/sec difference over the total number of CBR packetsPause time 10 s received gave the average end-to-end delay for theMaximum speed 20 m/s received packets. This metric describes the packetSimulation time 1000 s delivery time: the lower the end-to-end delay theNumber of nodes 10,20,30,40,50 (attack better the application performance. result on only 50 nodes)Area 1000 m * 1000 mPropagation Model Free spaceMaximum Malicious 10/4nodesMovement Model Random waypointTypes of attack Black-hole5 RESULTS AND DISCUSSIONSTo ensure a high-quality product, diagrams andlettering MUST be either computer-drafted ordrawn using India ink.5.1 DSR and AODV are with 50 nodes without attack. Fig: 3 Average end-to-end delay Of DSR (Series 1) and AODV (series 2)Considering the mobility of nodes and the network 1489 | P a g e
  4. 4. Pradish Dadhania, Sachin Patel / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 3, Issue 1, January -February 2013, pp.1487-14915.1.3 Packet Delivery Ratio Throughput(kbps) vs number of malicious nodesaPacket delivery ratio can be calculated as the ratio 120between the number of data packets that are sent bythe source and the number of data packets that are 100received by the sink. Graph shows as the number of 80node increase it gets better because probability ofpath breakage decrease. 60 40 20 0 2 4 6 8 10 MALICIOUS MALICIOUS MALICIOUS MALICIOUS MALICIOUS Fig 5 throughput Of DSR and AODV with malicious nodes Fig shows the graph for end to end delay. Graph shows as the number of malicious node increase it gets decrease output. In the DSR End to end delay is decreased and in AODV that increased. number of malicious nodes Vs End to EndFig 4 Packet Delivery Ratio of DSR (Series 1) and delay AODV (series 2) 1005.2 DSR and AODV are with 50 nodes with 80 attack. 60It is observed from the Fig that, the impact of the 40Black hole attack to the Networks throughput. The 20throughput of the network also decreases due toblack hole effect as compared to without the effect 0of black hole attack. We vary the speed of the node 2 4 6 8 10 MALICIOUS MALICIOUS MALICIOUS MALICIOUS MALICIOUSand take the result to the different node speed. Throughput(kbps) vs number of malicious number of malicious nodes Vs End to End nodes delay 120 100 100 80 80 60 60 40 40 20 20 0 0 2 4 6 8 10 2 4 6 8 10 MALICIOUS MALICIOUS MALICIOUS MALICIOUS MALICIOUS MALICIOUS MALICIOUS MALICIOUS MALICIOUS MALICIOUS Fig 6 Average end-to-end delay Of DSR and AODV with malicious nodes Fig shows the graph plotted between the numbers of malicious nodes and the packet delivery ratio. It 1490 | P a g e
  5. 5. Pradish Dadhania, Sachin Patel / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 3, Issue 1, January -February 2013, pp.1487-1491is inferred from the graph that since there is no Hole attack than DSR.mechanism available to detect malicious node inDSR, the packet delivery ratio decreases as the no REFERENCESof malicious node increases. [1] Varsha Patidar,Rakesh Verma, “Risk Mitigation of Black Hole Attack for Aodv Packet Delivery Ratio in % vs number of Routing Protocol,” IOSRJCE, Volume 3, malicious nodes Issue 3,July-Aug. 1 [2] N. Bhalaji, A. R. Sivaramkrishnan, Sinchan Banerjee, V. Sundar, and A. 0.8 Shanmugam. , “Trust Enhanced Dynamic 0.6 Source Routing Protocol,” World Academy of Science, Engineering and 0.4 Technology 49 2009. [3] C. Y. Lin, M. Wu, J. A. Bloom, I. J. Cox, 0.2 and M. Miller, “Rotation, scale, and translation resilient public watermarking 0 for images,” IEEE Trans. Image Process., 2 4 6 8 10 vol. 10, no. 5, pp. 767-782, May 2001. MALICIOUS MALICIOUS MALICIOUS MALICIOUS MALICIOUS [4] Nor Surayati Mohamad Usop, Azizol Abdullah, Ahmad Faisal Amri Abidin, Packet Delivery Ratio in % vs number of „„Performance Evaluation of AODV, malicious nodes DSDV & DSR Routing Protocol in Grid Environment‟‟, IJCSNS International 1 Journal of Computer Science and Network Security, VOL.9 No.7, July 2009 Q. Feng 0.8 , Z. Cai, J. Yang, and X. Hu . "A 0.6 Performance Comparison of the Ad Hoc Network Protocols." Second International 0.4 Workshop on Computer Science and Engineering. 2009. 0.2 [5] E. Gerhards-Padilla, N. Aschenbruch, and P. Martini. "Detecting Black Hole Attacks 0 in Tactical MANETs using Topology 2 4 6 8 10 Graphs." 32nd IEEE Conference on Local MALICIOUS MALICIOUS MALICIOUS MALICIOUS MALICIOUS Computer Networks. 2007. [6] Rajiv Misra and C.R.Manda, Fig 7: Packet Delivery Ratio Of DSR and AODV „„Performance Comparison of with malicious nodes. AODV/DSR On-demand Routing Protocols for Ad Hoc Networks in6 CONCLUSION Constrained Situation‟‟, Indian Institute of Technology, Kharagpur (India).We have done simulation for 50 nodes ad hoc [7] Vahid Nazari Talooki & Koorush Ziarati,network. As a traffic parameter we have used „„Performance Comparison of RoutingConstant Bit Rate. As far as mobility concern we Protocols For Mobile Ad Hoc Networks‟‟,are using Random Way Point Model. We have Dept. of Computer Science andtaken average of 10 simulations to make result Engineering, School of Engineering,more appropriate. We analyze both protocols in Shiraz Universityterms of throughput, Delay, Routing Overhead and [8] marjan kuchaki rafsanjani, ali movaghar,Packet Delivery Ratio. and faroukh koroupi, „„InvestigatingWe have analyzed during Black Hole Attack, based Intrusion Detection Systems in MANETon the number of attacker, the Packet Delivery and Comparing IDSs for DetectingRatio is high or low. If the number of them Misbehaving Nodes‟‟, World Academy ofincreases, the Packet Delivery Ratio is low, Science, Engineering and Technology 44because we are dropping data packets. As far as 2008.throughput concern, as number of malicious nodeincrease our throughput decreases because nodesare not able to find path towards destination andthat causes dropping. Based on our research andanalysis of simulation result we draw theconclusion that AODV is more vulnerable to Black 1491 | P a g e