Localization with witnesses

Abstract

Localization protocols enable an entity (called the verifier) to determine the physical location of another entity (called the prover), even if the prover maliciously advertises a false location or tries to corrupt the verifier's time measurements by time-shifting its responses. Unfortunately, the correctness of such protocols is critically dependent on the verifier's ability to make high-resolution time measurements and on the prover's ability and trustworthiness to send its response by the mandated time. To address these problems, we propose the idea of incorporating passive witnesses into the localization protocol. All witnesses monitor the same bilateral packet exchange between the prover and lead verifier and later report their respective inter-packet time measurements to the lead verifier for further processing. We show how the extra information provided by the witnesses can eliminate the threat of response-time shifting by a malicious prover. We also pose the question, how can we combine multiple localization observations to a single localization estimate? While analyzing that, we observe that the localization estimate is sensitive to the relative position of the prover among the verifiers © 2007 Springer.