Efficient routing attack detection and mitigation approach for mobile adhoc networks using node statistics

Abstract

The performance of the mobile adhoc network is highly dependent newlineon the energy parameter of the mobile nodes. Firstly a trust model is newlinepresented based on the computation of different parameters to improve the newlineperformance of intrusion detection in mobile adhoc network. The trust model newlinecomputes the direct trust between different agents. Similarly the method newlinecomputes the indirect trust which is propagated by the direct trust which is newlinecomputed between third party. Finally the method computes the newlinerecommendation trust between different agents to perform intrusion detection. newlineUsing all these measures the method performs intrusion detection on the newlinemobile adhoc network.In the second approach, the behavior of the user in accessing newlinedifferent web services is analyzed. The method computes the trust on the user newlinein accessing the web service and data service. The method classifies the user newlinequery on two different level; one on the application layer and another on the newlinedatabase level. Using both of the method, the trust weight is computed to newlineperform intrusion detection.Finally to improve the performance of the routing attack detection, newlinean efficient location aware traffic approximation scheme has been designed in newlinethis research work. The method monitors the network and stores the logs newlineabout the data packet being forwarded. At each time when a node has a packet newlineit performs neighbor discovery and route discovery. Then for each route the newlinemethod compute the trust weight and each intermediate node has been newlineverified for its location factor. Based on trust weight computed the method newlineselects an efficient forwarding neighbor to send the data packet. newline newline