Abstract:The structured sparse model incorporates dependencies among coefficients, and exploiting intrablock correlations significantly enhances recovery performance. The reconstruction performance of existing block sparse recovery algorithms relies on prior knowledge of blockwise sparse coefficients, including the number of blocks, boundaries, and sizes, which severely restricts the applicability of algorithms in scenarios like radar reconstruction. This article proposes a regularized block sparse Bayesian algorithm with structured prior learning, which involves a two-stage processing procedure. The first stage includes the acquisition of space–temporal hierarchical prior information. It introduces a regularization penalty term in the cost function to constrain and optimize hyperparameters, thereby promoting block correlations among neighboring coefficients. In the second stage, refined recovery is executed for nonzero blocks using a model with multiple measurement vectors (MMVs) constructed from space–temporal dimensional data. This model captures temporal correlations among multiple snapshots and blocks correlations within the prior coefficient structure, enabling joint parameter reconstruction across multiple domains. Simulation results demonstrate that compared with the conventional block sparse Bayesian learning (BSBL) algorithm, the proposed approach exhibits significant advantages in reconstruction performance under an unknown block sparse model.

Block-Sparse Signal Recovery Based on Adaptive Matching Pursuit Via Spike and Slab Prior

Efficient Recovery of Structured Sparse Signals Via Approximate Message Passing with Structured Spike and Slab Prior

Adaptive matching pursuit for sparse signal recovery

Approximate Message Passing with Nearest Neighbor Sparsity Pattern Learning

Unitary Approximate Message Passing for Sparse Bayesian Learning

Sparse Bayesian Learning Approach for Discrete Signal Reconstruction

Adaptive One-Bit Quantisation Via Approximate Message Passing with Nearest Neighbour Sparsity Pattern Learning

Block Sparse Recovery Via Mixed L 2 / L 1 Minimization

A Multilayer Joint Regularized Block Sparse Bayesian Algorithm With Space–Time Structured Prior Learning Function

Robust Recovery of Structured Sparse Signals with Uncertain Sensing Matrix: A Turbo-VBI Approach.

Simultaneous Block-Sparse Signal Recovery Using Pattern-Coupled Sparse Bayesian Learning

Block Sparse Bayesian Recovery with Correlated LSM Prior

Improved Sparse Signal Recovery Method Based on Matching Pursuit Revised MAP Estimation

Efficient Recovery of Block Sparse Signals by an Improved Algorithm of Block-StOMP

Extension of SBL Algorithms for the Recovery of Block Sparse Signals With Intra-Block Correlation

Exploiting Prior Information in Block Sparse Signals

A Sparsity Preestimated Adaptive Matching Pursuit Algorithm

An improved algorithm based on sparsity adaptive matching pursuit

Block-Sparse Signal Recovery via General Total Variation Regularized Sparse Bayesian Learning

Variance State Propagation for Structured Sparse Bayesian Learning

Recovery of Block Sparse Signals by a Block Version of StOMP