Adaptive Widely Linear Reduced-Rank Beamforming Based on Joint Iterative Optimization

Abstract

We propose a reduced-rank beamformer based on the rank-Joint Iterative Optimization (JIO) of the modified Widely Linear Constrained Minimum Variance (WLCMV) problem for non-circular signals. The novel WLCMV-JIO scheme takes advantage of both the Widely Linear (WL) processing and the reduced-rank concept, outperforming its linear counterpart as well as the full-rank WL beamformer. We develop an augmented recursive least squares algorithm and present an improved structured version with a much more efficient implementation. It is shown that the improved adaptive scheme achieves the best convergence performance among all the considered methods with a low computational complexity.Index Terms-Adaptive beamforming, linear constrained minimum variance, non-circular data, recursive least squares algorithms, reduced-rank methods, widely linear processing.

Design of Compressed Sensing System With Probability-Based Prior Information

“…As M ≪ N, (1) is an underdetermined problem which has infinite solutions. In order to find an unique mapping between the signal x and the measurement y, the constraint of sparsity on x can be utilized [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], Corresponding author: Sheng Li (email: shengli@zjut.edu.cn). [37], [38], [39], [40], [41], [42], [43], [44], [45].…”

Compressed Randomized Utv Decompositions for Low-rank Matrix Approximations in Data Science

“…Such compact representations which retain the key features of a high-dimensional matrix provide a significant reduction in memory requirements, and more importantly, computational costs when the latter scales, e.g., according to a high-degree polynomial, with the dimensionality. Matrices with low-rank structures have found many applications in background subtraction [1,2], system identification [3], IP network anomaly detection [4,5], latent variable graphical modeling [6], subspace clustering [7,8] and sensor and multichannel signal processing [9,10,11,12,13,14,15], [16,17,18,19,20,21,22,23,24,25,26,27]. .…”

Widely Linear Sphere Decoder in MIMO Systems by Exploiting the Conjugate Symmetry of Linearly Modulated Signals

“…[2] investigated the beamformers for extracting an unknown signal from non-circular interferences. To improve the convergence performance of widely linear beamformers, [4] proposed a widely linear joint-iterativeoptimization (WL JIO) beamformer based on the widely linear constrained minimum variance (WLCMV) criterion, which took the advantages of both WLP and reduced-rank techniques. [5] proposed widely linear minimum variance distortionless response (MVDR) beamformers for unknown second-order (SO) non-circularity coefficient situations (e.g.…”