Trusted DSR Routing Protocol to withstand against Gray Hole attack in MANET using Q-Learning

Abstract

-Routing is one of the required tasks in MANET to exchange packets among the mobile nodes that is accomplished by routing protocols. Traditionally routing protocols are designed to cope with routing operation but in practice they may be affected by misbehaving nodes in the form of attacks. Gray hole attack is one of the notable attack since they are launched internally and cannot be identified easily. It aims to disrupt the normal routing operation and try to minimize or collapse the overall network performance. Therefore, detecting Gray node and secure routing in MANET environment is always a challenging problem due to its distinct characteristics. In this paper, we have proposed a trusted DSR Routing protocol to detect the Gray hole attack over DSR routing protocol. Here we avoid the Gray hole nodes by calculating an overall trust of each node based on metrics such as direct and indirect trusts. In addition Q-learning algorithm is used to enrich the proposed work by assigning reward. This model is the extension of our earlier work [1] that overcomes the black hole attack over AODV Routing protocol.