Model-based decentralized Bayesian algorithm for distributed compressed sensing

作者：

Highlights：

• We proposed wavelet-based multi-scale model in Bayesian DCS.

• We used Bessel-K-form distribution in our algorithm, named WBDCS-BKF.

• WBDCS-WBKF uses both the tree structure and neighborhood relation of wavelet coefficients.

• VB inference procedure is used to derive posterior distributions.

摘要

•We proposed wavelet-based multi-scale model in Bayesian DCS.•We used Bessel-K-form distribution in our algorithm, named WBDCS-BKF.•WBDCS-WBKF uses both the tree structure and neighborhood relation of wavelet coefficients.•VB inference procedure is used to derive posterior distributions.

论文关键词：Distributed compressive sensing,Joint sparsity,Wavelet-tree structure,Bessel K-form,Variational Bayesian inference

论文评审过程：Received 17 September 2020, Revised 29 December 2020, Accepted 9 February 2021, Available online 17 March 2021, Version of Record 23 March 2021.

论文官网地址：https://doi.org/10.1016/j.image.2021.116212