Detecting Selfish Node over the Active Path using Neighbor Analysis based Technique

Sumiti, Dr. Sumit Mittal

Abstract: - In this work, a nearest neighbor analysis has been performed to detect the selfish nodes in the active path and to generate a secure path. The existing AODV protocol is modified and a new bit is taken to define the trustful status. If status is 1, then node is valid, otherwise the node is selfish node and no communication is performed over that node. As the communication is performed, each node is analyzed by its neighboring nodes and builds a trust table. The reply status is 0 (by default) as the successful replied is received by a node, the value in the table changed to 1. Now the protocol checks the shared table and identifies the reply status. If the reply status is greater than the threshold value, the node is taken as the valid node and communication over that node is performed. This work has been implemented using NS-2.29 simulator and results shows that this technique is able to detect almost 90 % selfish nodes in the active path.

Keywords: Manet, Selfish node attack, Nearest neighbor Analysis, Ns-2, AODV