Concentrative Intelligent Reflecting Surface Aided Computational Imaging via Fast Block Sparse Bayesian Learning
Junjie Yao, Zhaoyang Zhang, Xiaodan Shao, Chongwen Huang, Caijun Zhong, Xiaoming Chen
Abstract
Recently, millimeter wave (mmWave) imaging has received widespread attention. However, due to its nonlinearity and ill-posedness, it is challenging to reconstruct the precise electromagnetic properties of unknown targets from the measured scattered fields. In this paper, a new concentrative intelligent reflecting surface (IRS) aided computational imaging scheme is proposed. In the scheme, by dividing the region of imaging (ROI) into pixels, the imaging process is transformed into a compressed sensing problem. This paper proposes a fast block sparse Bayesian learning (BSBL) algorithm, which exploits the block sparsity of the reflection vector of ROI, and reduces the computational complexity through the generalized approximate message passing (GAMP) algorithm. Finally, the simulation results validate the performance advantages of the proposed algorithm and the efficiency of IRS in the imaging process.