CREDND: A Novel Secure Neighbor Discovery Algorithm for Wormhole Attack

Abstract

As nodes’ characteristics that they are self-governed and resource-limited, wireless sensor networks (WSNs) face potential threats due to various attacks, among which the most threatening attack is wormhole attack. Wormhole attack severely imperils WSNs and is difficult to be detected, for it causes incorrect routing by private tunnels and damages to WSNs in terms of data leakage, data dropping, and delayed delivery. However, the existing solutions are based on additional hardware, incur high communication overhead, or fail to give consideration to all types of wormholes. In this paper, we propose CREDND, a protocol for creating a Credible Neighbor Discovery against wormholes in WSN, which can detect not only external wormholes through the hop difference between the own exclusive neighbors but also internal wormholes through enabling the common neighbor nodes as witnesses to monitor whether the authentication packets are forwarded by malicious nodes. CREDND is a simple, localized protocol and needs no special hardware, localization, or synchronization, but it improves the ability of wormhole defense. The simulation results are provided, showing that CREDND outperforms in wormhole detection than other same types of solutions.