Trust Model for Secure Routing in Wireless Sensor Network using AI Technique

This paper proposes a Recommendation Filtering based Trust Model (RFTM) for securing a route between source to sink and filter out dishonest malicious nodes. RFTM filters the dishonest recommendations which are given by malicious neighboring nodes. It is a challenging problem due to the attacks like ballot and bad mouthing attack. This paper also explains the problem of attacks caused by malicious node while propagating trust information in the existing trust models. RFTM uses one of the artificial intelligence technique Dempster Shafer theory and deviation test scheme to avoid false recommendations from trust evaluation. The test results show their efficiency in improving throughput, packet delivery ratio of the network. The nodes are chosen in the test procedure randomly to check the performance of the filtering procedure against neighboring nodes.