PLS initialized sequential estimator for target localization using AOA measurements

Target localization using AOA measurements has attracted substantial attention for several decades. Traditional algorithms regard the target position as a non-random parameter and employ estimators like least squares (LS) or maximum likelihood (ML) to estimate the target location. In this paper, we propose a new framework for target localization using AOA measurements. The idea of this framework is to treat the unknown position as a random vector and then use the linear minimum mean square error (LMMSE) criterion to obtain an estimator that sequentially fuses the AOA measurements from multiple sensors. The key difficulty of this criterion is how to determine the prior first two moments of the unknown location. This is tackled by pseudo-linear least squares (PLS), which is verified to be perfectly credible through three credibility measures. Extensive numerical examples show that the PLS initialized sequential estimator outperforms the existing PLS and its root-mean-square error (RMSE) is close to the Cramer-Rao lower bound (CRLB) in most cases.