Possibility and Necessity Measures to Enhance Reliability and Cooperation in MANETS Anoop J. Sahoo 1* and Md. Amir Khusru Akhtar 2 1 Infosys Limited Chennai, India; anoop.jyoti88@gmail.com 2 Department of Computer Engineering, Cambridge Institute of Technology, Ranchi, Jharkhand, India; akru2008@gmail.com Abstract Ensuring security in Mobile Ad hoc Network (MANET) is a major concern. The majority of attacks can be prevented by secure routing protocol, but non-cooperation still suffers from a serious drawback which is the total strength of nodes in a network. A lot of solution to enforce cooperation has been proposed in the literature which isolate nodes directly on the basis of lesser reputation values but reduces the total strength of nodes in a network. Finally, the lesser strength of nodes degrades the performance and reliability of mobile ad hoc networks. To enforce cooperation in MANET this work presents the Perfect Evidence (PE) model which uses reputation value to obtain the possibility and necessity measures and isolate a node having perfect evidence. It involves nesting of focal element to know the perfect evidence. The proposed model enhances performance of the network because it eliminates only one misbehaving node in each turn, which maintains the total strength of nodes in a network. In addition to that it enhances cooperation in network because it identifies and isolates the most misbehaving node and warns other misbehaved nodes. The warning message suggests other misbehaving nodes to cooperate in network activities otherwise at any turn it will be isolated from the network. Experimental result shows the efficiency of the model. Keywords: Evidence, Focal Element, Necessity Measure, Possibility Measure, Reputation Value *Author for correspondence 1. Introduction A MANET is an independent body of mobile nodes connected by wireless links. ese networks work in a standalone fashion in which each node has to cooperate in network activities. But, in order to save its valuable resources nodes drop packets of others which degrades the efficiency of packet transfer, increases the packet delivery time, enhances the packet loss rate and creates network partitioning. e packet dropping attack or mis- behavior is further classified into selfish and malicious as discussed by Anusas-Amornkul 1 . In selfish misbehavior nodes drop packets of others for its honest causes such as battery life and bandwidth. In spite of that we have some malicious causes of data dropping attacks such as wormhole and blackhole. e other reason for data drop- ping attacks are network congestion, jamming and burst channel errors due to interference, fading etc. In this work, a node which drops packets of others is called a selfish node or misbehaving node. is chapter presents the Perfect Evidence (PE) model which uses reputation value to obtain the possibility and necessity measures and isolate a node having perfect evi- dence. It involves nesting of focal element to know the perfect evidence. e proposed model involves one or more expert nodes which are accountable for all major computations, analysis and isolation of misbehaving nodes. We have used reputation values to analyse and iso- late a node in each turn. In each turn we identify a single misbehaving node and warn other misbehaving nodes to cooperate in network activities. e identified misbehav- ing node is isolated from the routing paths. Our proposed model isolates only one node in each turns to maintain the total strength of nodes in a network which enhance Indian Journal of Science and Technology, Vol 7(3), 312–317, March 2014 ISSN (Print) : 0974-6846 ISSN (Online) : 0974-5645