Localization of sparse and coherent sources by orthogonal least squares

Area of Science:

• Signal Processing
• Array Signal Processing
• Computational Electromagnetics

Background:

• Estimating signal covariance is crucial for analyzing correlated sources, often found with distributed sources or reflections.
• Existing methods face computational complexity challenges, limiting their application in high-dimensional scenarios.

Purpose of the Study:

• To propose an efficient method for joint source localization and signal covariance estimation.
• To overcome the computational complexity associated with traditional covariance matrix estimation techniques.

Main Methods:

• Covariance Matrix Fitting by Orthogonal Least Squares (CMF-OLS).
• A greedy dictionary-based approach utilizing the orthogonal least squares algorithm.
• Exploiting orthogonal least squares to reduce computational load.

Main Results:

• The proposed method significantly reduces computational complexity compared to existing techniques.
• Achieves higher accuracy in source correlation matrix estimation.
• Demonstrates capability in handling high-dimensional problems and exploring extensive source position spaces.
• Successfully locates and identifies physical and mirror sources in experimental settings.

Conclusions:

• The CMF-OLS method offers an efficient and accurate solution for joint source localization and covariance estimation.
• Its reduced complexity enables analysis of complex scenarios with fine spatial discretization.
• Experimental validation confirms its effectiveness with correlated sources and reflectors.