Compressive Sensing-Based Detector Design for SM-OFDM Massive MIMO High Speed Train Systems

Abstract

In this paper, spatial modulation (SM)-orthogonal frequency division multiplexing is introduced to massive multi-input multi-output high speed train systems to reduce the complexity of the receiver. First, the mathematic model of signal detection in frequency domain is formed as a compressive sensing (CS) problem. Based on this model, a novel detector, which exploits the unique group sparsity of the signal, is proposed to improve the accuracy of the nonzero elements localization. On the basis of the channel matrix decomposition, the detector combines the CS algorithms and the iteration process together to overcome the performance deterioration brought by inter-carrier interference. Moreover, theoretical analysis about the performance gain guarantee of our proposed detector is conducted and the complexity of the proposed detector is compared with the related works. Furthermore, we also extend our proposed detector into the special scenario that the train passes the overlapped area of the adjacent two cells and receives the superimposed signal. In order to deal with the inter-cell interference, a generalized SM framework is adopted to model the signal detection. Simulation results verify that the proposed detectors have an apparent performance gain compared to the conventional CS schemes with the same complexity order.