Array-Based Geolocation Exploiting Sensor Motion for CubeSat Applications

Small satellites are of increasing interest in applications that previously required large satellite buses. In this paper, we examine the use of a small satellite (under 1m in each dimension) for array-based geolocation of targets on the Earth's surface. The small satellite size precludes the possibility of arrays with a large number of elements, both because of the small aperture that can be supported and because the electronics of the RF chain would not fit in the satellite volume and power constraints. We develop an algorithm that uses direction information from different locations as the satellite sweeps out its orbit, based on arrays with as few as three antennas. A separate method determines ground target location using information from a single earth-based location. These different estimates can be fused in a minimum-variance combination. We study the variance of the position estimates as a function of the number of satellite positions and samples used to estimate the direction vectors at each position. As shown via extensive simulations, fusing multiple estimates with the single-location estimate decreases the standard deviation of the measurement error by up to three orders of magnitude. The method is also shown to be generally insensitive to Doppler shifts.