CS reconstruction in MIMO channel using square complex orthogonal STB codes

Abstract

Compressed sensing (CS) is the process of signal reconstruction at a rate far below the Nyquist sampling rate. Sometimes, CS measurements need transmission over radio mobile channel for bandwidth preserving and energy efficient system design and/or application specific system design such as wireless body area network (WBAN), remote surveillance system etc. To this aim, this work proposes an integrated system design for quality improvement in CS reconstructed images over radio mobile channel modelled as Rayleigh flat fading. The design of a non-zero-entry Single Symbol Decodable Square Complex Orthogonal, Space Time Block Codes (STBC) for 16 antennas is proposed first. This multi-channel facility is then exploited for transmission of CS measurements, which at the receiver side, undergo Robbins–Monro (RM) method, a non-parametric stochastic approach for predicting the unobserved spaces from the observed ones through recursive filtering. Similar study is also done using a relaxed convex optimization of the basis pursuit algorithm, a parametric based approach. Extensive simulation results highlight the efficacy of the proposed CS and STBC (CS–STBC) scheme for the quality improvement on the reconstructed images as the variation in the number of antennas, change in the number of measurements and variation on signal-to-noise ratio. It also highlights significant performance gain in CS reconstruction over radio mobile channel using STBC framework.