Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.



Published on

  • Be the first to comment

  • Be the first to like this


  1. 1. Harsh Lohiya, Rajnish Choubey, Roopali Soni / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 3, Issue 1, January -February 2013, pp.1029-1035 Performance Improving and Securing Routing Backup Protocol for MANET from Selective Forwarding Attack Harsh Lohiya Rajnish Choubey Roopali Soni M. Tech. IV Sem , Dept. of C.S.E. Asst. Prof. , Dept. of C.S.E. Asst. Prof. , Dept. of C.S.E. Oriental College of Technology Oriental College of Technology Oriental College of Technology Bhopal (M.P) , India Bhopal (M.P) , India Bhopal (M.P) , IndiaAbstract- An ad hoc mobile wireless network the route has been modified, then, theconsists of a number of wireless mobile nodes attacker/intruder would perform different types ofthat are capable of communicating with each attacksother without the use of a network infrastructure such as Selective forwarding Attack ,Flooding,or any centralized administration. Each device in Neglect and Greed etc.a MANET is free to move independently in any In our approach , we have been secured thedirection, and will therefore change its links to network from selective forwarding attack. Theother devices frequently. MANET faces various selective forwarding Attack was first described byproblems related to their securities and various Karlof and Wagner [2]. This attack is sometimestypes of attacks create problem in network’s data called Gray Hole attack. In a simple form oftransmission. Selective forwarding attack is one selective forwarding attack, malicious nodes try toof them which is harmful attacks against mobile stop the packets in the network by refusing toadhoc network and capable to disturb the whole forward or drop the messages passing through communication. The various prevention There are different forms of selective forwardingtechniques against selective forwarding attack is attack. In one form of the selective forwardingoverwhelming. In this paper , we present an attack, the malicious node can selectively drops thealgorithm to defend against selective forwarding packets coming from a particular node or a group ofattacks based on AODV routing protocol which nodes. This behaviour causes a DOS attack for thatprovide secure data transmission or forward the particular node or a group of node. They alsodata safely, and detect the selective forwarding behave like a Blackhole in which it refuses toattack. In first phase, we judge the trust value of forward every packet. The malicious node mayeach node to select a secure path for message forward the messages to the wrong path, creatingforwarding to detect the malicious nodes which unfaithful routing information in the network.are suspected to launch selective forwarding Another form of selective forwarding attack isattack. We also present simulation results with called Neglect and Greed. In this form, theperformance evaluation of the proposed subverted node arbitrarily neglecting to route somealgorithm. To the best of our knowledge, this is messages [1]. It can still participate in lower levelthe first paper to present an algorithm for protocols and may even acknowledge reception ofdefending selective forwarding attacks in data to the sender but it drops messages randomly.MANET. Such a node is neglectful. When it also gives excessive priority to its own messages it is alsoKeywords- MANET; Selective Forwarding greedy. Moreover, another variance of selectiveAttack; Black hole attack ;AODV ; Attack; forwarding attack is to delay packets passing through them, creating the confused routingI. INTRODUCTION information between nodes. MANET is described as a self-configurable Selective forwardingand rapidly deployable wireless network. The attack is one of the harmful attacks against mobileabsence of centralized management makes each ad-hoc network and capable to disturb the wholewireless node in MANET to perform routing to its network communication. The various preventionneighbours in order to maintain the connectivity and techniques against selective forwarding attack isthe network stability. Therefore, the routing protocol overwhelming. In this paper , we present anmust ensure both connectivity and security algorithm to defend against selective forwardingto achieve the network stability. Unfortunately, the attacks based on AODV routing protocol whichwidely used routing protocols perform their provide secure data transmission or forward the dataalgorithms over MANET routing protocols assume safely, and detect the selective forwarding attack. Inthat all the nodes are trusted [19]. If the routing first phase, we judge the trust value of each node toinformation has been fabricated and the direction of select a secure path for message forwarding to 1029 | P a g e
  2. 2. Harsh Lohiya, Rajnish Choubey, Roopali Soni / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 3, Issue 1, January -February 2013, pp.1029-1035detect the malicious nodes which are suspected to S. Marti et al. [3] proposed a reputation-launch selective forwarding attack. To the best of based scheme in which two modules (i.e. watchdogour knowledge, this is the first paper to present an and pathrater) are added on at each node. Watchdogalgorithm for defending selective forwarding module maintains a buffer of recently sent orattacks in MANET. forwarded data packets. Buffer is cleared only when This paper is structured as the following: Section watchdog overhears the same packet being2 explains the related work which had done forwarded by the next hop node over the mediumpreviously. Section 3 introduces the proposed work and if a data packet remains in the buffer too long,with complete explaination of our algorithm. the next hop neighbor is suspected to beSection 4 displays the simulation results and misbehaving. Based on watchdog‟s suspicion,performance evaluation. And Section 5 concludes Pathrater module maintains a rating for every otherthe paper and shows the future work. node in the network and calculates a path metric by averaging the node ratings in the path and then II. RELATED WORK chooses the best path. Main advantage of this To prevent routing misbehavior or scheme is that it can detect misbehavior at theselfishness in MANETs, various solutions have been forwarding level as well as in link level. But it mightproposed previously which can be roughly classified not detect misbehavior in presence of ambiguous[6] as: collisions, receiver collisions, limited transmission power, false misbehavior and partial dropping.A. Credit Based Scheme: A credit-based approach, Sonja Buchegger et al. [7] proposedon the other hand, uses the concept of virtual CONFIDANT protocol which is based on selectivecurrency. Nodes pay virtual money for services altruism and Utilitarianism. In CONFIDENT, trust(networking resources) that they get from other relationships and routing decisions are based onnodes, and similarly, get paid for providing services experienced, observed, or reported routing andto other nodes. forwarding behavior of other nodes. It consists of four modules: The Monitor, the Reputation System, S Zhong et al. [6], proposed a credit based the Path Manager, and the Trust Manager. Eachscheme each node maintains receipts for messages node monitors the behavior of its next-hop nodewhich are received and forwarded. When the nodes continuously and if a suspicious activity is detected,get a connection to a credit clearance service, they information of the suspicion is passed to thereport those credits, and based on the decision taken Reputation System. The Reputation System changesby the CCS the nodes need to pay or they may be the rating of the suspected node which depends onrewarded with real money. Since this uses an how significant and how frequent the activity is andexternal party for the payment, it may not be useful if rating of a node becomes less than certainfor all scenarios. threshold, control is passed to the Path Manager. Buttyan and Hubaux et al [5] , used the To prevent selfishness in MANET, K.concept of beans (nuggets) as payments for packet Balakrishnan et al. [8] proposed a TWOACKforwarding. They proposed two models: packet scheme which can be implemented as an add-on topurse model and packet trade model. In packet purse any source routing protocol. Instead of detectingmodel, beans are loaded into the packet before it is particular misbehaving node, TWOACK schemesent. The sender puts a certain number of beans on detects misbehaving link and then seeks to alleviatethe data packet to be sent. Each intermediate node the problem of routing misbehavior by notifying theearns beans in return for forwarding the packet. If routing protocol to avoid them in future routes. It isthe packet exhausts, the beans in it drops before done by sending back a TWOACK packet onreaching its destination. In the packet trade model, successful reception of every data packet, which iseach intermediate node buys the packet from the assigned a fixed route of two hops in the directionprevious node for some nuggets. Thus, each opposite to that of data packets.intermediate node earns some beans for providing Basic drawback of this scheme includes it cannotthe forwarding service and the overall cost of distinguish exactly which particular node issending the packet is borne by the destination. misbehaving node. Sometime well behaving nodes became part of misbehaving link and therefore canB. Reputation Based Scheme: not be further used the network. Thus a lot of well In a reputation-based approach, nodes behaved node may be avoided by network which(either individually or collectively) detect, and then results in losing of well behaved routes.declare another node to be misbehaving. This K. Vijaya et al. [9] proposed anotherdeclaration is then propagated throughout the acknowledgement based scheme similar tonetwork, leading to the misbehaving node being TWOACK scheme, which is also integrated on topavoided in all future routes. of any source routing protocols. This scheme detects the misbehaving link, eliminate it and choose the 1030 | P a g e
  3. 3. Harsh Lohiya, Rajnish Choubey, Roopali Soni / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 3, Issue 1, January -February 2013, pp.1029-1035other path for transmitting the data. The main idea is trust value of each node to select a secure path forto send 2ACK packet which is assigned a fixed message forwarding to detect the malicious nodesroute of two hops back in the opposite direction of which are suspected to launch selective forwardingthe data traffic route and to reduce the additional attack. Different from the multi-path routing whichrouting overhead, a fraction of the data packets will only defends the selective forwarding attack; ourbe acknowledged via a 2ACK packet. This fraction method may find the malicious termed as Rack and by varying the Rack, Our protocol ensures that multicast data isoverhead due to 2ACK packets can be dynamically delivered from the source to the members of thetuned. This scheme also consists of multicasting multicast group, even in the presence of attackers, asmethod by which sender can broadcast information long as the group members are reachable throughof misbehaving nodes so that other nodes can avoid non-adversarial path.path containing misbehaving nodes and take another Here an authentication framework is used topath for the data transmission. Although routing eliminate outside adversaries and ensure that onlyoverhead caused by transmission of authorized nodes perform certain operations (onlyacknowledgement packets is minimized but this tree nodes can perform tree operations and onlyscheme also suffers to detect the particular group nodes can connect to the correspondingmisbehaving node. multicast tree). Srdjan et al. [10] proposed a two-fold approach Our protocol mitigates attacks that try to preventfor detection and isolation of nodes that drops data a node from establishing a route to the multicast treepackets. First approach attempts to detect the both in route request and route reply.misbehavior of nodes and will identify the malicious Our protocol involves following steps :activity in network. It is done by sending an ACK (1) Trust key computingpacket by each intermediate node to its source node (2) Secure node authenticationfor confirming the successful reception of data (3) Secure route discovery across the node.packets. If the source node does not get ACK packet Select a node to destinationby intermediate nodes then source node send again Check selected node in fresh_route cacheits packet for destination after a specific time. If If yes thensame activity was observed again then source node Route is confirmedbroadcast a packet to declare the malicious activity Elsein the network. Other approach identifies exactly Select another new secured nodewhich intermediate node is doing malicious activity. End ifIt is done by monitoring the intermediate nodes of (4) Backup node setup route by the nodes near to active path which (5) Route maintenance across the node.lies in their transmission range and by the nodes A. Trust Key Computingwhich are on the active route. Since monitoringnodes are in promiscuous mode and are in the There are lots of protocols have been devised totransmission range of intermediate nodes of active secure ad hoc mobile wireless protocols usingroute, they can receive all the packets sent along the cryptography. These cryptographic protocols workactive route. Monitoring nodes count the number of under the presence of a central authority.packet coming into and going out of the nodes of A new parameter weight value named TLv can beactive route. Each monitoring node maintain a list of used to choose the best path which ensuressent and dropped packets and when number of reliability of the path by calculating the trust valuedropped packets by a particular node exceeds certain of the neighbor nodes and that value can be stored inthreshold, the monitoring node in that range declares a priority table. Each time a node sends a RREQthat node as misbehaving node and broadcast this either when it determines that it should be a part of ainformation. Upon receiving broadcast packet all multicast group, and it is not already a member ofneighboring nodes will cancel their transmission to that group, or when it has a message to send to thethat particular node and enter it into the list of multicast group but does not have a route to thatmisbehaving nodes. Main disadvantage of this group. An intermediate node after receiving ascheme includes the overhead due to transmissions RREQ packet updates its path in the routing tableof acknowledgement packets by every intermediate and add the TLv value of its link and forward it tonode to the source and working of all nodes in the next node.promiscuous mode. To calculate the trust value a new trust policy has been introduced in link and network layer toIII. PROPOSED WORK calculate a key which can be used to determine the In order to avoid the selective forwarding reliability of neighbor node, where the keyattack, we proposes a scheme of secure data calculation involves dynamic assignment oftransmission which can forward the data safely, and weights. The policy resides in route entry trustdetect the selective forwarding attack. We judge the computing part, operates independently and 1031 | P a g e
  4. 4. Harsh Lohiya, Rajnish Choubey, Roopali Soni / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 3, Issue 1, January -February 2013, pp.1029-1035maintains its individual perspective of trusthierarchy. An entity gathers information about the data andcontrol packet of its neighboring node and overhearsdata from the events like whether a packet or controlmessage is dumped and not retransmitted. Based onthis, every node will maintain some values in a tablefor its entire neighboring node.B. Secure Node Authentication The authentication framework prevents untrustednodes to be part of a multicast tree or join amulticast tree. Each node forwards RREQIRREPonly when the node from which RREQIRREP is Figure 1. Mobile ad-hoc Network with sourcereceived must be a trust node. Node maintains a node 1neighbor list, when neighbors calculated trust valueis less than the threshold NEIGH UNSECURE, then C. Secure route discovery across the nodemarked it as not credible and unset enable flag in When source node requires the route tomulticast routing table. destination, source enters the route discovery phase Every source will maintain a table which contains and checks whether adequate “fresh” routes todestination host, next hop, interface and the average destinations are already available in the Fresh_routetrust level value for the existing paths available as in cache. If some “fresh” routes to destination inTable I. These fields can be updated based on the Fresh_route cache are found, source runs Routereceived RREP messages. An alternative route confirm process. Otherwise, source runs newdiscovery can be initiated for a significantly low secured route discovery process to find a securedTSTv value for a particular route considering that new route to the destination node.route non-reliable even if there is no link breakage. Source node broadcasts RD_request to nearby Here we see that how above trust nodes; RD_request includes a sequence numbervalue table maintain in source. Let‟s take the field to distinguish the route discovery process fromMANET ,shown in figure 1 in which eight movable others , a route content field for node address alongnodes are present. Node 1 is a source node . in the path from S to D and the trust level of thewhich the trust value table is to be stored. Trust source. After the intermediate node receivesvalue table in node 1 is as follows: RD_request from an upstream node X, it inserts its TABLE 1 address into the route content field of the TRUST VALUE TABLE RD_request only if it is in the same trust level of the source by confirming the trust key and then sendsDestination Next Interface TSTv this modified RD_request to its neighboring nodes Hop (excluding the upstream node X). The RD_request3 3 2 4 cache of the intermediate node also records the information, including the sequence number of the2 2 8 3 RD_request and which neighboring nodes are sent only if the request is not duplicated. Otherwise, the8 4 5 2 duplicated request is discarded. If a “fresh” route is available from source to the5 2 6 5 destination in the Fresh_route cache, the source node S adds the secured fresh route from S to D to6 4 8 7 the RC_request and then transmits RC_request along this route. When it receives the RC_request, an intermediate node checks its Fresh_route cache to determine whether any other fresh route to D is included. If a “fresh” route is available, the node copies RC_request and puts the route information in the route content field of the RC_request before transmitting the RC_request along this fresh route. If no “fresh” route is available, RC_request is transmitted downstream according to its route content field. Eventually, after D receives the RC_request, RD_reply is sent back to S, and S sends packets through this original route. 1032 | P a g e
  5. 5. Harsh Lohiya, Rajnish Choubey, Roopali Soni / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 3, Issue 1, January -February 2013, pp.1029-1035D. Backup node setup phase RE_request to announce all backup nodes that this When RD_request or RC_confirm reaches the data transmission process is ending. The RE_requestdestination D, it may gather many secured routes packet contains the sequence number of thiswith in a period „TC‟. The nodes of those routes transmission process for distinguishing it from otherwhich D received are compared pair wise from process. When the backup node receivesbeginning to end to find whether any two paths have RE_request, it will also save remnant secureda section in common. The final node, excluding backup routes from backup_route cache indestination D, in such a section is the “backup Fresh_route cache.node”. A subset of backup nodes can be gatheredfrom any two secured routes. Then, all the subsets of IV. SIMULATION AND RESULTSbackup nodes are joined and the BS_ packet that We implemented our protocol in NS2includes simulator, a popular network simulator for MANET each backup node and the partial path from the to investigate the performance of our proposedbackup node to the destination node are generated. schemes. In the experiments, 50 mobile nodes moveThe destination node then uses BS_packet to within a rectangular area of 670m × 670m. At theseparately setup the backup_route cache of those same time, we set up the maximum speed as 5m/s,backup nodes, where the BS_packet contains the and the pause time as 20 seconds before each nodesequence number of this secured routing process, can move to its next destination. During the processthe address of a back up node under the path from of communication, the traffic is generated overthe backup node to the destination. The backup UDP. For each node, the transmission range is set tonodes store the partial paths from the backup node 250m without fading effect. Additionally, someto the destination node in their backup_route cache nodes are not willing to cooperate for routing andafter they receive the BS_packet. data delivery but every malicious node actsE. Route maintenance across the node independently. When a link fails, a node cannot continue to TABLE IITransmit . The node sends an error message, SIMULATION PARAMETERSlink_fail_message, to an upstream node along the Parameter type Parameter valuereverse current route. This message is used toannounce the back up node alone in the route to Simulation time 300sreplace the secured backup route. The alert message Simulation terrain 670 . 670mwill not be passed by an upstream node until the Number of nodes 50message is returned to a backup node. When the Mobility model Random waypointbackup node receives the message of link failure, Mobility 0-20m/sthe secured backup route from backup_route cache Temperature 290Kis fetched to replace the route behind the backupnode, and the source node S is informed to change Path loss model Two-raythe route. Thus, the node S sends the packets along Radio frequency 2.4 GHzthe new secured route. If backup_route cache Channel bandwidth 2 Mbps includes no other secured backup route, then the MAC protocol 802.11node has lost the identity of the backup node. Under Transmission range 250 msuch circumstances, no backupnode exists. Thesource node will receive the link_failure_message CBR data sessions 10and re-enters the route discovery phase to establish a CBR data rate 4 packets per secondsecured new route to the destination. After the Packet size 512 bytesdestination node replies with a path back to thesource as the current route for sending data packets, The simulation environment is a 670 · 670 squaresome secured backup routes are established and meters, where 50 nodes are randomly distributed.stored in backup nodes. If the current route is still Node pairs are randomly selected to generatealive, the situation that any node along the secured CBR/UDP traffic. Channel bandwidth is 2 Mbps.backup route moves will not influence the The path loss model is Two-Ray Ground Model.communication of the current route. If the secured The CBR data packet size is 512 bytes and thecurrent route is broken and replaced by a back up packet rate is 4 packets per second. The detailedroute, it can still work even though a section of this simulation parameters are listed in Table II. Thebackup route has failed. That is because the link random waypoint mobility model is used in ourwhich failed will be detected and an alert message simulation. Each node randomly selects a position,will be sent to find another back up node. When S and moves toward that location with a randomlydoes not have the route to D, S will store the usable generated speed between the minimum and theroute into the Fresh_route cache and broadcast maximum speed, which is 0 and 20 m/s, 1033 | P a g e
  6. 6. Harsh Lohiya, Rajnish Choubey, Roopali Soni / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 3, Issue 1, January -February 2013, pp.1029-1035respectively. Once it reaches that position, itbecomes stationary for a predefined pause time.After that pause time, it selects another position andrepeats the process as mentioned above. We changethe pause time to simulate different mobility rates.The pause time is set from 0 to 300 s. When thepause time is equal to 300 s, it means all nodes staystill during the simulation.A. Simulation result Because AODV simply drops data packets whena route becomes disconnected, the packet deliveryratio of AODV is the worst one among the otherschemes. The total number of packets received by thedestination using our routing protocol is higher thanthe ad hoc backup routing protocol.B. Performance Evaluation Graph 1 showing the comparision inperformances of Mobile Ad-hoc networks whensecurity is provided in MANET and when securityisnot provided in MANET. When we have appliedsecurity in MANET , the number of bits received isalso increases as compared to without securitynetwork. According to our throughput graph, weconclude that our approach improves the dataefficiency of MANET and also solves the problemof selective forwarding attack which drops the datapackets when this approach is not applied in GRAPH 1. Throughput graph showing theMANET. In above graph , blue line shows the performances of MANET with security and withoutthroughput of network when security is applied and security .orange line shows the throughput of network whensecurity isnot applied. V. CONCLUSION AND FUTURE WORK MANET is a self-configurable and rapidly deployable wireless network. The absence of centralized management makes each wireless node in MANET to perform routing to its neighbours in order to maintain the connectivity and the network stability. Security is one of the major issues in MANETs. In order to avoid the selective forwarding attack Our protocol enhances the routing protocol that solves most of its security flaws, prevents and detects attack. First, we incorporate a trust key computing model which can be used to choose the best path and ensure reliability of the path by calculating the trust value of the neighbor nodes. Then we perform secure node authentication in which authenticated node can only be participated. We proposed a scheme of secure data transmission which can forward the data safely, and detect the selective forwarding attack. We judge the trust value of each node to select a secure path for message forwarding to detect the malicious nodes which are suspected to launch selective forwarding attack. 1034 | P a g e
  7. 7. Harsh Lohiya, Rajnish Choubey, Roopali Soni / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 3, Issue 1, January -February 2013, pp.1029-1035 In future we plan to detect most of the attacks, Telecommunication Technologies ,which are common to ad hoc network routing December 2009, pp. 576-578.protocols. We are also planning to work on power [12] Muhammad Zeshan, Shoab A. Khan, Ahmadoptimization of the network. Raza Cheema and Attique Ahmed, "Adding Security against Packet Dropping Attack inREFERENCES Mobile Ad Hoc Networks," in 2008International Seminar on Future[1] G. Lavanya, C.Kumar and A. Rex Macedo Information Technology and Management Arokiaraj, “Secured Backup Routing Engineering, November 2008, pp. 568-572. Protocol for Ad hoc Networks” IEEE 2010, [13] Songbai Lu, Longxuan Li, Kwok-Yan Lam, p-45-50 Lingyan Jia, SAODV: A MANET Routing[2] Chris Karlof , David Wagner, “Secure routing Protocol that can Withstand Black Hole in wireless networks: attacks and Attack, International Conference on countermeasures”. Computational Intelligence and Security,[3] S. Marti, T. Giuli, K. Lai, and M. Baker, IEEE 2009, p 421-425 “Mitigating Routing Misbehavior in Mobile [14] FengHe, Kuan Hao, Hao Ma, S-MAODV:A Ad Hoc Networks,” in Proc. of the Sixth Trust Key Computing Based Secure Annual International Conference onMobile Multicast Ad-hoc On Demand Vector Computing and Networking (MobiCom Routing Protocol, IEEE 2008, p434-438 2000), August 2000, pp. 255-265. [15] Elahe Sheklabadi, Mehdi Berenjkoub, ,An[4 Yih-Chun Hu, Adrian Perrig, David B. Anonymous Secure Routing Protocol for Johnson,Rushing Attacks and Defense in Mobile Ad Hoc Networks, IEEE 2011, p 142- Wireless Ad Hoc Network, WiSe, ACM, 147 2003, p 30-40 [16] Intrusion Detection System in wireless Ad-[5] L. Buttyan and J.P. Hubaux, “Enforcing hoc Networks Based on Mobile Agent Service Availability in Mobile Ad-Hoc Technology, IEEE 2010, p 470- 475 WANs,” Proc. MobiHoc, Aug. 2000. [17] R. S. Mangrulkar, Dr. Mohammad Atique,[6] S. Zhong, J. Chen, and Y. Yang, “Sprite: A Trust Based Secured Adhoc on Demand Simple, Cheat Proof, Credit- Based System Distance Vector Routing Protocol for Mobile for Mobile Ad-Hoc Networks,” in Proc. of Adhoc Network, IEEE 2010 IEEE INFOCOM‟03, March 2003, pp. 1987- [18] Imad Aad, JeanPierre Hubaux, and Edward 1997. W. Knightly, Denial of Service Resilience in[7] Sonja Buchegger Jean-Yves Le Boudec, Ad Hoc Networks, MobiCom‟ IEEE 2004, p- “Performance Analysis of the CONFIDANT 202-215 Protocol Cooperation Of Nodes Fairness In [19] Maha Abdelhaq, Sami Serhan, Raed Dynamic Ad-hoc NeTworks" in Proc. Alsaqour and Rosilah Hassan „A Local IEEE/ACM Workshop Mobile Ad Hoc Netw. Intrusion Detection Routing Security over Comput. (MobiHoc 2002), June 2002, pp. MANET‟ International Conference on 226-236. Electrical Engineering and Informatics 17-19[8] K. Balakrishnan, D. Jing and V. K. Varshney, July 2011, Bandung, Indonesia "TWOACK: Preventing Selfishness in Mobile Ad hoc Networks," in Proc. of Wireless Communications and Networking Conference (WCNC‟05), vol. 4, March 2005, pp. 2137-2142[9] K.Vijaya “Secure 2Ack Routing Protocol In Mobile Ad Hoc Networks,” TENCON 2008, IEEE Region 10 Conference, November 2008, pp. 1-7. [10] Srdjan C apkun, Jean-Pierre Hubaux, and Levente Buttya´n, Mobility Help security in Ad Hoc Networks, MobiHoc, ACM 2003, p 46-56[11] S. Usha, S. Radha, "Co-operative Approach to Detect Misbehaving Nodes in MANET Using Multi-hop Acknowledgement Scheme," in 2009 International Conference on Advances in Computing, Control, and 1035 | P a g e