PERFORMANCE ANALYSIS OF THE MAXIMUM A-POSTERIORI PROBABILITY DIRECTION FINDING

In this paper, the mean square error performance of the maximum a posteriori (MAP) probability direction finding by sensor array in terms of its Cramer-Rao lower bound (CRLB) is analyzed. Based on the principle of Bayesian estimator, a log posteriori probability function is formed when the a priori knowledge of location is given. The Fisher information matrix (FIM) is found accordingly. It shows that the CRLB of the MAP estimator is much lower than that of maximum likelihood technique, especially when the element SNR is low and/or the number of snapshots is small. In addition, the CRLB remains at a relatively low level in terms of the variance of DOA of sources. It also shows that the location variance dominates the behavior of the MAP direction finder when locations of sources are Gaussian distributed.