A Cross-layer Method for Identifying and Isolating the Blackhole Nodes in Vehicular Ad-hoc Networks

Abstract

ABSTRACT Vehicular Ad-hoc Network (VANET) is a set of intelligent vehicles that communicate without a fixed infrastructure. The communication between each source/destination pair is done by using routing protocols. On-demand multipath distance vector (AOMDV) is one of the most known ad-hoc multipath routing protocols used in VANETs. The decentralized nature of VANET makes this type of network vulnerable to various attacks, such as blackhole attack. In such attack, the malicious vehicle aims to make the communication unavailable. To achieve this goal, the malicious vehicle persuades the source to send its data packets through it because it has the fresher route toward the destination. This is done by forging routing information. After receiving the data packets, the malicious vehicle deletes them instead of forwarding them to their intended destinations. This paper introduces a new Cross-Layer method (CRAOMDV) where information is shared between MAC and network layers to detect and ignore the malicious vehicles in VANETs. Our experiments used the simulator NS2 and SUMO for the generation and simulation of real mobility scenarios. The evaluation results demonstrate the efficiency of CRAOMDV compared to AOMDV under blackhole attack in terms of improving the packet delivery and reducing the average end-to-delay and the routing overhead.