Performance Bounds of a Single LEO Satellite Providing Geolocation of an RF Emitter

The recursive Posterior Cramer Rao Bound (rPCRB) is computed for a Low Earth Orbit (LEO) satellite providing geolocation estimates of an RF emitter. This bound takes no assumptions on the geometry of the situation, allows geolocation parameters to vary over time, and can be meaning-fully compared with nonlinear estimation algorithm performance. A stationary emitter is assumed, it’s only velocity coming from the earth’s rotation, which is modeled and estimated. The measurements are instantaneous frequencies corrupted with AWGN noise. Previous relevant literature uses simplified geometric assumptions and the traditional Cramer Rao Bound (CRB), producing reasonable results only in benign and select scenarios. Here, more realistic performance bounds are given and the potential of geolocating emitters in ’search and rescue’ scenarios is demonstrated. The performance bounds are computed for various scenarios and compared with an Unscented Kalman Filter using estimated instantaneous frequencies from synthesized data.