DOA Estimation Based on Weighted l₁-norm Sparse Representation for Low SNR Scenarios

Area of Science:

• Signal Processing
• Array Signal Processing

Background:

• Direction of Arrival (DOA) estimation is crucial for wireless communication systems.
• Low signal-to-noise ratio (SNR) conditions pose significant challenges for traditional DOA algorithms.
• Existing l1-norm-based singular value decomposition (L1-SVD) methods struggle with accuracy in sparse or noisy environments.

Purpose of the Study:

• To propose a novel weighted l1-norm approach within the L1-SVD framework.
• To enhance the accuracy and resolution of DOA estimation, particularly in low SNR scenarios.
• To suppress spurious peaks and improve signal sparsity for more robust estimation.

Main Methods:

• A weighted l1-norm is integrated into the L1-SVD algorithm.
• A weighted matrix is derived by optimizing subspace orthogonality.
• The weighted l1-norm serves as the objective function to promote signal sparsity, approximating the l0-norm.

Main Results:

• Simulated results for Orthogonal Frequency Division Multiplexing (OFDM) signals show narrower main lobes and lower side lobes.
• The proposed algorithm demonstrates improved resolution and accuracy with fewer snapshots and reduced sensitivity to misestimation.
• Experimental results confirm superior performance over existing methods in low SNR conditions, with narrower main lobes and lower side lobes.

Conclusions:

• The weighted l1-norm L1-SVD algorithm significantly improves DOA estimation accuracy and resolution in low SNR environments.
• The method effectively suppresses spurious peaks and enhances signal sparsity.
• The algorithm shows practical applicability for DOA estimation of Unmanned Aerial Vehicles (UAVs) and pseudo base stations.