Emitter Localization and Visualization (ELVIS): A Backward Ray Tracing Algorithm for Locating Emitters

Abstract

We present ELVIS, a new approach for localizing a first responder (e.g., firefighter) inside a building. We assume that the first responder emits RF energy, which undergoes multiple reflections with the walls, ceilings and floors of the building. There are K receivers, each of which receives ray(s) from the first responder. Each receiver estimates the AOA (angle of arrival), TOA (time of arrival) and power of each ray. We assume that the receivers know the blueprint of the building and the electromagnetic characteristics of the construction material used in the building. We show that, based on this information, the receivers can localize the first responder to a high degree of accuracy by applying ELVIS, which is based on backward ray tracing. We have evaluated the performance of ELVIS, using both single and multiple receivers, under a variety of channel and propagation conditions. We find that, the location prediction error depends mostly on the angular resolution of the receiver antennas. One advantage of ELVIS over other RF-based localization methods is that, if AOA is estimated with high degree of accuracy, a single receiver would be sufficient to do localization in 3D. In this case one could localize 100% of the Bell Labs, Crawford Hill building within 9 cm. With degraded accuracy, including fading and noise effects, plusmn10 degree angular error, and plusmn10 ns temporal error, one could localize 80% of the building within 10 m using multiple receivers.