Super-resolution direction-of-arrival estimation based on generalized eigendecomposition and sparse reconstruction...

Area of Science:

• Signal Processing
• Array Signal Processing
• Electromagnetics

Background:

• Super-resolution sparse recovery methods struggle with strong near-field interference.
• Existing solutions like dictionary augmentation or pre-filtering have limitations, causing basis coherence or signal distortion.

Purpose of the Study:

• To develop a robust preprocessing method for direction-of-arrival estimation in the presence of strong near-field interference.
• To overcome limitations of current sparse recovery techniques in challenging environments.

Main Methods:

• A novel preprocessing method based on generalized eigenvalue decomposition is proposed.
• The method employs a matrix filtering concept to maximize the power ratio between desired and interference subspaces.
• It extends optimal beamforming to a matrix filtering framework, balancing interference suppression and target preservation.

Main Results:

• The proposed method significantly improves super-resolution direction-of-arrival estimation performance under strong near-field interference.
• It demonstrates effectiveness even with limited snapshots and low signal-to-noise ratios.
• Simulation and experimental results validate the method's superiority over existing approaches.

Conclusions:

• The generalized eigenvalue decomposition preprocessing method offers a computationally efficient and effective solution for near-field interference mitigation.
• It enables robust super-resolution direction-of-arrival estimation in practical scenarios.
• The approach provides a balanced trade-off, preserving target signals while suppressing interference.