Emitter geolocation using low-accuracy direction-finding sensors

Abstract

This paper examines the concept of replacing a few expensive high-accuracy radio direction-finding (DF) sensors operating in a stand-off baseline with many low-accuracy DF sensors deployed on existing military vehicles located throughout an area of interest. A formula is presented to estimate the geolocation accuracy that could be achieved for a given number of sensors with a specified DF accuracy. Monte Carlo and virtual simulation of sensors with varying DF accuracy is used to assess the accuracy and reliability of the geolocation estimates that could be achieved. Geolocation results are computed using Stansfield's method and a technique developed at DRDC Ottawa, called the Discrete Probability Density (DPD) method, and compared to the theoretical location accuracy limit predicted by the Cramer-Rao Lower Bound. The DPD method is shown to provide more accurate and more reliable geolocation estimates than Stansfield's method when incorporating large quantities of low-accuracy DF data. This suggests that the DPD method could be used with several less-expensive low-accuracy DF sensors to provide improved emitter geolocation capability compared to the conventional deployment of a few high-accuracy DF sensors.