Enhanced Detection of Malicious Nodes in Ad-hoc on Demand Distance Vector Protocol using Cross Layer Design

Abstract

A Mobile Ad-Hoc Network (MANET) is a popular type of wireless network that is formed by a collection of mobile nodes. Each node in such network has the capability to communicate with its neighbors and non-neighbors through a wireless medium designed to accommodate the properties of a self-organized environment without protection against any inside or outside network attacks. In addition to, it does not have any central authority to ensure that any node in the network it is not malicious, which introduces security flaws to the networks. we propose security policy to detect malicious nodes over the Ad hoc On Demand Distance Vector (AODV) MANET routing protocol. To improve security in AODV routing protocol- which adds extra features to it and making path formation more secure- we are Using Cross-layer design shares knowledge about the state and conditions of one layer to the other layers. In this work, the link and routing layer are considered. In the security policy. The algorithm that is used in AODV protocol benefits from the in-formations feedback-coming from link to network layer- clarifying the number of requests sent from the nodes is shown to indicate whether the node is malicious or not. The improvement of the system performance considers the achieved result of this research using the mat-lab simulator. It evaluates the AODV security performance metrics in term of end-to-end delay which has average of values equal 7.9140, the overall improvement increasing in the bit rate and Packet delivery ratio, and Throughput with average of values are 6.2595, 0, 0016 and 0, 9313 respectively.