Impact of Coherent Sources on Direction Finding Using Maximum Likelihood Algorithm

To investigate the impact of coherent sources on direction finding using the Maximum Likelihood (ML) algorithm, we configured two coherent signals and conducted detailed experiments and analysis. The simulation implementation involves generating correlated signal sources through covariance matrix manipulation and applying ML estimation using optimization techniques like Newton-Raphson iteration or gradient descent. Through these experiments employing spatial spectrum estimation and peak detection algorithms, we gained deeper insights into how coherent sources affect direction finding results, thereby providing guidance for improving direction finding algorithms. Furthermore, we examined how different characteristics of coherent sources (such as correlation coefficient and phase relationship) influence direction finding accuracy through parameter sweeping tests, and proposed corresponding solutions including spatial smoothing preprocessing and forward/backward averaging techniques. These research findings, implemented through MATLAB's signal processing toolbox and custom optimization functions, contribute to the development and application of coherent source direction finding technology.