Directions-of-Arrival Estimation Through Bayesian Compressive Sensing Strategies

• Published in: IEEE transactions on antennas and propagation Vol. 61; no. 7; pp. 3828 - 3838
• Main Authors: Carlin, Matteo, Rocca, Paolo, Oliveri, Giacomo, Viani, Federico, Massa, Andrea
• Format: Journal Article
• Language: English
• Published: New York, NY IEEE 01-07-2013; Institute of Electrical and Electronics Engineers; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Antennas; Applied sciences; Arrays; Bayes methods; Bayesian compressive sensing (BCS); Compressed sensing; Direction-of-arrival estimation; direction-of-arrival estimation (DoA); Electromagnetism; Engineering Sciences; Estimation; Exact sciences and technology; linear array; Radiocommunications; Sensors; Signal and Image processing; sparse representation; Studies; Telecommunications; Telecommunications and information theory; Vectors; Performance evaluation; Parameter estimation; Direction-of-arrival estimation; State of the art; Bayesian compressive sensing (BCS); Correlation matrix; Measurement sensor; Signal estimation; Numerical method; direction-of-arrival estimation (DoA); Addressing; Implementation; linear array; Narrow band signal; Output voltage; Sparse representation; Linear antenna arrays; Compressed sensing
• ISSN: 0018-926X; 1558-2221
• DOI: 10.1109/TAP.2013.2256093

The estimation of the directions of arrival (DoAs) of narrow-band signals impinging on a linear antenna array is addressed within the Bayesian compressive sensing (BCS) framework. Unlike several state-of-the-art approaches, the voltages at the output of the receiving sensors are directly used to determine the DoAs of the signals thus avoiding the computation of the correlation matrix. Towards this end, the estimation problem is properly formulated to enforce the sparsity of the solution in the linear relationships between output voltages (i.e., the problem data) and the unknown DoAs. Customized implementations exploiting the measurements collected at a unique time instant (single-snapshot) and multiple time instants (multiple-snapshots) are presented and discussed. The effectiveness of the proposed approaches is assessed through an extensive numerical analysis addressing different scenarios, signal configurations, and noise conditions. Comparisons with state-of-the-art methods are reported, as well.