Résumé: We propose a distributed mechanism, Dis-VoW, to detect wormhole attacks in under-water sensor networks. In Dis-VoW, every sensor reconstructs local network layout using multi-dimensional scaling. It detects the wormholes by visualising the distortions in edge lengths and angles among neighbouring sensors. The contributions include: • Dis-VoW does not depend on any special hardware • it provides a localised wormhole detection mechanism adapting to network topology changes • it integrates techniques from social science and scientific visualisation to attack network security problems. The simulation results show that Dis-VoW can detect most of the fake neighbour connections without introducing many false alarms.

Langue: Anglais