Source localization by summing multiple correlator outputs

A technique for multisensor, passive localization is examined, and a comparison of localization performance with the Cramer-Rao lower bound is presented. The technique tests a weighted sum of correlations obtained at time delays corresponding to a hypothesized location; a grid search over the hypothesized location; a grid search over the hypothesized locations is used to find the location that best explains the observed measurements. The Cramer-Rao lower bound on localization accuracy is obtained for sensors in arbitrary positions with multipath and sensor-position inaccuracy. The bound is validated by comparison with standard results for more restrictive cases. Simulation results are presented for a restricted number of scenarios, including cases of path decorrelation and mismodeling of the number of propagation paths present