Wormholes no more? Localized Wormhole Detection and Prevention in Wireless Networks

A number of protocols have been proposed to date to defend against wormhole attacks in wireless networks by adopting synchronized clocks, positioning devices, or directional antennas. In this work, we introduce a novel approach for detecting wormhole attacks. The proposed algorithm is completely localized and works by looking for simple evidence that no attack is taking place, using only connectivity information as implied by the underlying communication graph, and total absence of coordination. Unlike many existing techniques, it does not use any specialized hardware, making it extremely useful for real-world scenarios. Most importantly, however, the algorithm can always prevent wormholes, irrespective of the density of the network, while its efficiency is not affected even by frequent connectivity changes. We also provide an analytical evaluation of the algorithm’s correctness along with an implementation on real sensor devices that demonstrates its efficiency in terms of memory requirements and processing overhead.