Maximum likelihood and Cramer-Rao lower bound estimators for (nonlinear) bearing only passive target tracking

Cramer-Rao lower bound (CRLB) is a powerful fool in assessing the performance of any estimation algorithm. Nardone, et al. (1984) developed maximum likelihood estimator (MLE) for passive target tracking using batch processing. In this paper, this batch processing is converted into sequential processing so that if is useful for the above real time application using bearings only measurements. Adaptively, the weightage of each measurement is computed in terms of its variance and is used along with the measurement making the estimate a generalized one. Instead of assuming some arbitrary values, pseudo linear estimator outputs are used for the initialization of MLE. The algorithm is tested in Monte Carlo simulation and its results are compared with that of CRLB estimator. From the results, it is observed that this algorithm is also an effective approach for the bearing only passive target tracking.