Localization under Adversary Misdirection.

摘要

Wireless sensor networks are particularly versatile for performing a wide range of detection and estimation tasks for the identification, localization, and tracking of unknown phenomenon. The shared nature of the wireless medium presents a number of design challenges as the sensors transmitted data can be compromised both by natural artifacts such as noise and channel fading and by the intentional corruption due to an adversary. While much work has examined how noise and channel conditions impact estimation accuracy, much less attention has been given to the problem of corruption due to an intelligent adversary. This work considers the problem of estimating emitter location in the presence of an intelligent adversary. The localization problem is considered from two different viewpoints. We focus on the fundamental behavior of adversary-network strategies at the physical layer. Examination of an adversary at the physical layer is a rich space for discovering fundamental behavioral insights into security strategies for sensor networks. Strategies that both the network and the adversary should employ are developed, to see which side really has the upper hand. The adversary seeks to degrade and redirect the network while the networks goal is to mitigate effect of the adversary. These physical layer adversary attacks and defense strategies can be considered the last line of defense when higher-layer techniques fail.