DESIGN AND ANALYSIS OF ADHOC ON DEMAND DISTANCE VECTOR ROUTING (AODV) IN WIRELESS SENSOR NETWORK USING NEURAL NETWORK

Recently, Wireless sensor networks became a hot research topic among researchers due to their flexibility and independence of network infrastructures, such as base stations. Due to unique characteristics, such as dynamic network topology, limited bandwidth, and limited battery power, routing in WSN is a particularly challenging task compared to a conventional. And it has experienced serious security problems due to their particular characteristics. Wireless communication can endure interferences or malicious interceptions; whereas, multi-hop communication assumes that each node will perform properly its functions to support network services. Further, self-organization increases the complexity of security management operations as access control, node authentication, secure routing and cryptographic key distribution. Network early work in WSN research has mainly focused on developing an efficient routing mechanism in such a highly dynamic and resource-constrained network. At present, several efficient routing protocols have been proposed for WSN. Most of these protocols have assumed a trusted and cooperative environment. However, in the presence of malicious nodes, the networks are vulnerable to various kinds of attacks. AODV protocol gives high performance gain. In WSN, routing attacks are particularly serious. Aims and objectives of this thesis work is to design and implement NN protocol with SINK hole attack and prevent the system for threat using this hybridisation.