Near-Field Source Localization in Nonuniform Noise: An Efficient Symmetric Matrix Factorization-Based Approach.

Area of Science:

• Signal Processing
• Array Signal Processing
• Electromagnetics

Background:

• Near-field source localization is crucial for various applications.
• Existing methods often struggle with unknown nonuniform noise.
• Noise heterogeneity can significantly degrade localization accuracy.

Purpose of the Study:

• To develop a robust near-field source localization method for multiple narrowband signals.
• To address the challenge of unknown nonuniform noise with arbitrary diagonal covariance.
• To improve localization accuracy and robustness in complex noise environments.

Main Methods:

• Reformulated the problem as joint symmetric matrix factorization and noise variance estimation.
• Developed an iterative algorithm based on the block majorization-minimization principle.
• Derived the Cramér-Rao bound (CRB) for near-field localization under nonuniform noise.

Main Results:

• The proposed algorithm achieves high accuracy in near-field source localization.
• Demonstrated superior performance compared to state-of-the-art methods.
• The approach shows strong robustness against severe nonuniform noise.

Conclusions:

• The novel covariance-fitting approach effectively handles nonuniform noise.
• The developed iterative algorithm is computationally efficient.
• The method provides near-optimal performance, closely matching the derived CRB.