SLAT: Secure Localization with Attack Tolerance

Accurate and secure localization is essential to the correct operation of many applications of sensor networks. However, existing methods lack concrete security mechanisms or are not resilient to beacon node compromise. We address the limitations of present approaches in this paper through the Secure Localization with Attack Tolerance (SLAT) protocol. In SLAT, non-beacon nodes estimate their positions by calculating the intersection of multiple authenticated beacon messages. Message authentication prevents wholesale beacon location report forgeries. To combat compromised beacons, we develop and analyze a location reporting algorithm that ensures that compromised beacons have little ability to affect location estimates. Moreover, the degree to which a malicious location report affects an estimate is inversely proportional to its distance from the true location (as reported by the majority of properly operating beacons). We evaluate the protocol via simulation within a range of sensor networks and protocol parameters. Our results indicate that even large numbers of compromises only nominally affect location estimates. For example, we show that compromising 40 out of 200 nodes in a simulated environment only increases the average location estimate error from 3 meters to 5 meters. Such results apply across a wide range of networks, topologies, and hardware platforms.