Abstract

This paper presents the BLSML method, which combines Brown's Least Square Error (LSE) and Maximum Likelihood (ML) techniques for accurate target position estimation. The proposed method, named BLSML, offers a passive approach that leverages the strengths of both LSE and ML to improve the estimation accuracy. In the first stage, BLSML provides an initial estimated position using LSE, which takes advantage of the robustness and efficiency of the LSE algorithm. Then, in the second stage, BLSML performs optimization based on ML principles, refining the estimated position to enhance accuracy. We have utilized the Cramer-Rao Lower Bound (CRLB) on the estimator's covariance to develop an efficient estimator for position estimation. To assess the performance of the proposed BLSML method, comprehensive Monte Carlo simulations were conducted. Realistic noise and error models were incorporated to mimic real-world conditions and evaluate the method's robustness. The evaluation metrics included Root Mean Square Error (RMSE) for position estimation accuracy and Circular Error Probable (CEP) for error region analysis. The results demonstrated that BLSML achieved remarkable accuracy, with an RMSE below 0.43%, an average estimation error of 1.907% and the radius of the 90% CEP region of approximately 925 m for target position estimation within a 200-kilometer range.