A Single Satellite Geolocation Solution of an RF Emitter Using a Constrained Unscented Kalman Filter

Abstract

A real-time geolocation solution of a radio frequency (RF) emitter is presented using a constrained Unscented Kalman Filter (cUKF). Samples are processed on a single low earth orbit (LEO) satellite by a cUKF that imposes constraints by projecting the sigma points into the feasible solution space. The emitter is transmitting an RF signal from an unknown l0-cation with a known center frequency. The ease of imposing constraints within the UKF framework for single LEO satellite RF geolocation is demonstrated and results show reduced convergence time, high resiliency to noise, sub-kilometer geolocation accuracies, and the maintenance of stability. An accurate and real-time single satellite solution can be a valuable asset, as it could greatly reduce cost and computation time of existing multi-satellite systems. This paper shows the feasibility of such a solution and provides a framework to approach more realistic link budget, satellite GPS inaccuracies, and oscillator drift scenarios.