PROJECT TITLE :

Alternative to Extended Block Sparse Bayesian Learning and Its Relation to Pattern-Coupled Sparse Bayesian Learning - 2018

ABSTRACT:

We tend to consider the matter of recovering block sparse signals with unknown block partition and propose a higher different to the extended block sparse Bayesian learning (EBSBL). The underlying relationship between the proposed methodology EBSBL and pattern-coupled sparse Bayesian learning (PC-SBL) is explicitly revealed. The proposed technique adopts a cluster-structured previous for sparse coefficients, which encourages dependencies among neighboring coefficients by properly manipulating the hyperparameters of the neighborhood. Thanks to entanglement of the hyperparameters, a joint sparsity assumption is made to yield a suboptimal analytic answer. The different algorithm avoids high dictionary coherence in EBSBL, reduces the unknowns of EBSBL, and explains the effectiveness of EBSBL. The proposed algorithm additionally avoids the vulnerability of parameter choice in PC-SBL. Results of comprehensive simulations demonstrate that the proposed algorithm achieves performance that is shut to the simplest performance of PC-SBL. Additionally, it outperforms EBSBL and alternative recently reported algorithms significantly underneath noisy and low sampling eventualities.