Publication

Abstract : Compressed sensing based detection technique has emerged to be a promising solution for large-scale sporadic multi-user MIMO systems. However, their performance significantly degrades with a decrease in sparsity, increased constellations and imperfect channel estimation. In this paper, an improved iterative thresholding technique is proposed to resolve the above problems. Bit error rate performance and computational complexity characteristics for varying MIMO configurations are presented. Results substantiate that the proposed technique offers a gain of around 2 dB compared to conventional compressed sensing algorithm. The proposed technique seems to be a viable solution to improve the performance without compromising the complexity.