Design and Improve AODV Protocol for Congestion Avoidance in MANET Using Neural Network

Abstract- The mobile adhoc networks are the decentralized type of network in which mobile nodes can communicate with each other without the presence of central controller. Due to non-presence of central controller routing, security and quality of service are the three major issues of the network. The congestion is the issue of the mobile adhoc networks which reduce performance of the network in terms of various parameters. In this work, the improvement in the AODV protocol is done for the congestion avoidance in the network. The proposed technique is based on the back propagation algorithm in which error of each path is available from source to destination is calculated. The best path is selected which has minimum error or means that which has least chances of congestion in the network. The proposed and existing algorithms are implemented in NS2 and it has been analyzed that proposed technique performs well in terms of various parameters as compared to existing algorithm. Keywords: MANET, AODV, RREQ, RREP, Back Propagation, Ad-hoc.