Improved Estimation of Trilateration Distances for Indoor Wireless Intrusion Detection
Detecting wireless network intruders is challenging since logical addressing information may be spoofed and the attacker may be located anywhere within radio range. Accurate indoor geolocation provides a method by which the physical location of rogue wireless devices may be pinpointed whilst providing an additional option for location-based access control. Existing methods for geolocation using received signal strength (RSS) are imprecise, due to the multipath nature of indoor radio propagation and additional pathloss due to walls, and aim to minimise location estimate error. This paper presents an approach to indoor geolocation that improves measurements of RSS by averaging across multiple frequency channels and determining the occurrence of walls in the signal path. Experimental results demonstrate that the approach provides improved distance estimates for trilateration and thus aids intrusion detection for wireless networks.