A Probabilistic Energy-Efficient Approach for Monitoring and Detecting Malicious/Selfish Nodes in Mobile Ad-hoc Networks

MANETs are networks composed of mobile nodes connected using the wireless medium to communicate each other. The MANET nature makes it subject to various types of security leaks. External nodes could interfere with the communication, fair nodes could be compromised at any time by malicious agents. Another issue regards the limited amount of energy available to nodes, so their behavior could become selfish under some conditions. Many proposals detect these behaviors using a trust management system, but the activity of packet monitoring is energy-consuming. The main aim of this proposal is the reduction of the energy required by the monitoring system without lowering detection performance. Various simulations show that the achievement of this result is obtained through the application of a probabilistic approach, having also better performance in terms of time needed to detect misbehaviors.