Resilient secure localization and detection of colluding attackers in WSNs

There exists extensive work in wireless sensor networks (WSNs) on security measures that guarantee the correctness of the estimation of the position of a node despite attacks from adversaries. But very little work has investigated how colluding attackers can modify the behavior of known attacks or even create new ones. In this paper, we first present an attack model that allows three types of colluding attackers to threaten the secure localization process and/or the attacker detection process. We then describe a decentralized algorithm that is used to determine the position of a location-unknown sensor U despite the presence of colluding attackers (that can alter any type of information being exchanged in a WSN in order to form an attack jointly). Most importantly, the proposed algorithm allows U to detect such colluding attackers in its sensing range. Our simulation results show that in both a uniformly deployed WSN en