Efficient Reduced-Rank DOA Estimation Algorithms Using Alternating Low-Rank Decompositions

TL;DR

This paper introduces an alternating low-rank decomposition method with recursive least squares algorithms for efficient DOA estimation, demonstrating superior performance over existing techniques in large sensor arrays.

Contribution

It presents a novel ALRD approach with RLS-based algorithms for improved DOA estimation, especially in large arrays with correlated sources.

Findings

Algorithms outperform existing methods in simulations.

Effective for large sensor arrays with correlated sources.

Provides accurate DOA estimates with reduced computational complexity.

Abstract

In this work, we propose an alternating low-rank decomposition (ALRD) approach and novel subspace algorithms for direction-of-arrival (DOA) estimation. In the ALRD scheme, the decomposition matrix for rank reduction is composed of a set of basis vectors. A low-rank auxiliary parameter vector is then employed to compute the output power spectrum. Alternating optimization strategies based on recursive least squares (RLS), denoted as ALRD-RLS and modified ALRD-RLS (MARLD-RLS), are devised to compute the basis vectors and the auxiliary parameter vector. Simulations for large sensor arrays with both uncorrelated and correlated sources are presented, showing that the proposed algorithms are superior to existing techniques.