A reliable data delivery using trust management system based on node behaviour predication in MANET

An improvisation version of traditional wireless networks provide mobile ad hoc network (MANET), which is highly suitable for emergency needs. But, at the same time, due to its infrastructure less and resources limitation create many challenges in its performance. Mainly due to its dynamic behaviour it prones to high security concern, as the complete communication cycle depends on the independent nodes which voluntary join and leave. Several security studies reveal that trust management systems can be effective and attained low overhead. But, most of these studies used an automated decision algorithms based on some predefined trust scheme. This kind of scheme unable to distinguish the node intentional and unintentional errors at runtime, which impact the selected paths for the communication. So, it is very important to identify node behaviour and also provide node recovery scheme. In this paper, we propose a novel trust management system based on node behaviour predication algorithm to preserve high network stability and security for the reliable data delivery. The algorithm will predict and identify the node behaviour and distinguish between unintentional temporary errors and intentional malicious behaviour and compute a node overall trust. Experimental results show that, node behavior prediction and trust management system improvise the throughput ratio and network stability.