Optimal Sensor Placement for a Constellation of Multistatic Narrowband Pixelated Sensors

This paper presents a novel approach to selecting the location of a constellation of multistatic RF sensors to optimize system detection and tracking. The method works by computing the posterior Cramér-Rao lower bound on localization error as a function of sensor positions, and then selecting sensor locations to minimize the bound. One important contribution of this study is the derivation of the bound for a non-Gaussian, nonlinear pixelated sensor model which includes transmit and receive beam patterns, direct path energy, and target impulse response. A second contribution is a report on the results of two field collections which illustrate the efficacy of the method.