Wireless Intrusion Detection Using Tracking Verification

In this work we report on a wireless intrusion detection system based on the use of position tracking for location verification. Using optimal error bounds developed for extended Kalman filters as the comparison basis, we show how wireless signal strength measurements can be used to reliably detect intruders whose true position tracks are far removed from their claimed tracks. We show how the system is immune to malicious attacks that adjust transmitted signal strengths in attempts to circumvent the intrusion detection algorithms. We compare our results with earlier work that utilized static position information as a means for intrusion detection, showing how verification errors determined via tracking can be an order of magnitude smaller. Contrary to the static case, we also show how position verification via tracking is viable in the case where only one access point is within range of the mobile device. Our work outlines the theoretical underpinning of position verification via nonlinear tracking, in addition to providing detailed Monte Carlo simulations of the verification performance.