ON THE FEASIBILITY OF THE MULTIPATH FINGERPRINT METHOD FOR LOCATION FINDING IN URBAN ENVIROMENT

Abstract

The feasibility of the multipath fingerprint method wireless location finding in urban enviroments is examined using computational electromagnetics simulation. Fingerprints composed of time and angle of arrival data in urban environments are created using electromagnetic ray tracing. The fingerprints at specific locations within a simple four-by-four building model are studied to address issues including uniqueness of the fingerprints, repeatability due to environmental changes, and bandwidth limitations. A classifier based on template matching is also constructed using the simulation data from a 1-km city area of downtown Austin, Texas. Angle of Arrival (AOA) is found to be a stronger identifier than Time of Arrival (TOA). The classifier results demonstrate that good location-finding performance is achievable using both time of arrival and angle of arrival features.