Broadband Source DOA Estimation Using Forward-Backward Smoothing and Minimum Redundancy Linear Arrays

Broadband source DOA (Direction of Arrival) estimation represents a critical research area in array signal processing, with extensive applications in radar, sonar, and wireless communication systems. While traditional DOA estimation methods perform well under narrowband conditions, their performance deteriorates in broadband signal scenarios. This paper introduces an integrated approach combining forward-backward smoothing techniques with minimum redundancy linear arrays to enhance DOA estimation performance for broadband sources.

The forward-backward smoothing technique serves as an effective method for decorrelating coherent signals. Building upon conventional forward smoothing, it incorporates additional backward smoothing processing to further improve discrimination capability for coherent signals. For broadband signals, the frequency diversity effect enables forward-backward smoothing to better utilize multi-frequency point information, thereby enhancing estimation robustness. In code implementation, this typically involves constructing forward and backward covariance matrices through array data rearrangement, followed by eigenvalue decomposition for signal subspace estimation.

Minimum redundancy linear arrays constitute an optimized array configuration approach that achieves higher degrees of freedom with the same number of array elements by reducing redundant spacing between sensors. This configuration not only improves angular resolution but also reduces computational complexity. For broadband DOA estimation, the combination of minimum redundancy linear arrays with forward-backward smoothing can more effectively utilize array aperture, improving separation performance for multiple source signals. Algorithm implementation often requires optimized array geometry design using combinatorial mathematics principles, followed by wideband processing techniques like coherent signal subspace method (CSSM) or incoherent methods.

Simulation experiments validate the effectiveness of this methodology. Results demonstrate that compared to conventional approaches, the scheme based on forward-backward smoothing and minimum redundancy linear arrays achieves higher estimation accuracy and better robustness under low signal-to-noise ratio conditions and multiple source scenarios. This method proves particularly suitable for practical environments containing coherent or strongly correlated broadband signals. Key implementation steps include wideband focusing matrices construction, spatial smoothing covariance calculation, and MUSIC-like spectrum estimation algorithms.