Source Enumeration for High-Resolution Array Processing Using Improved Gerschgorin Radii Without Eigendecomposition

• Published in: IEEE transactions on signal processing Vol. 56; no. 12; pp. 5916 - 5925
• Main Authors: Lei Huang, Lei Huang, Teng Long, Teng Long, Shunjun Wu, Shunjun Wu
• Format: Journal Article
• Language: English
• Published: New York, NY IEEE 01.12.2008; Institute of Electrical and Electronics Engineers; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Additive noise; Applied sciences; Array signal processing; Arrays; Complexity; Computation; Computational efficiency; Covariance matrix; Detection, estimation, filtering, equalization, prediction; Deviation; Direction finding; eigenvalue decomposition (EVD); Eigenvalues; Eigenvalues and eigenfunctions; Enumeration; Exact sciences and technology; Gerschgorin radii; high resolution; Information, signal and communications theory; Mathematical models; minimum description length (MDL); Miscellaneous; multistage Wiener filter (MSWF); Noise level; Noise robustness; Radar signal processing; sensor array signal processing; Sensor arrays; Signal and communications theory; signal enumeration; Signal processing; Signal, noise; Telecommunications and information theory; White noise; Working environment noise; Gerschgorin Radii; High resolution; Multistage Wiener fiber (MSWF); Direction finding; Sensor array signal processing; Minimum description length (MDL); Eigenvalue decomposition (EVD); Signal enumeration; Performance evaluation; Direction-of-arrival estimation; Parameter estimation; sensor array signal processing; Array signal processing; signal enumeration; multistage Wiener filter (MSWF); Signal estimation; Radiolocalisation; Eigenvalue; Background noise; Wiener filter; eigenvalue decomposition (EVD); White noise; MDL criterion; Signal detection; Additive noise; Multistage method; Gerschgorin radii; Covariance matrix; Computational complexity; minimum description length (MDL); Signal processing; Numerical simulation; Sensor array
• ISSN: 1053-587X; 1941-0476
• DOI: 10.1109/TSP.2008.929331

Accurate detection of sources with low complexity is of considerable interest in practical applications of high-resolution array processing. This paper addresses a new computationally efficient method for source enumeration by using enhanced Gerschgorin radii without eigendecompositofion. The proposed method can calculate the Gerschgorin radii in a more efficient manner, in which the additive background noise can be efficiently suppressed and the computational complexity can be considerably reduced. Therefore, the method is more accurate and computationally attractive. Furthermore, the method does not rely on the eigenvalues of a covariance matrix or the signal/noise power, making it robust against deviations from the assumption of spatially white noise model. Numerical results are presented to demonstrate the performance of the method.