Efficient Compressive Sensing Detectors for Generalized Spatial Modulation Systems.
Lixia Xiao,Ping Yang,Yue Xiao,Shiwen Fan,Marco Di Renzo,Wei Xiang,Shaoqian Li
DOI: https://doi.org/10.1109/tvt.2016.2558205
IF: 6.8
2017-01-01
IEEE Transactions on Vehicular Technology
Abstract:Generalized spatial modulation (GSM) is a novel multiple-input-multiple-output (MIMO) technique, which relies on a sparse use of radio-frequency (RF) front ends at the transmitter. In this paper, low-complexity and compressive-sensing (CS)-based detectors for GSM systems are proposed. First, an extension of the normalized CS detector (E-NCS) based on the orthogonal matching pursuit (OMP) algorithm is proposed, which is shown to be suitable for large-scale GSM implementations, due to its low complexity. Furthermore, to mitigate the error floor effect of the E-NCS detector, two efficient CS (ECS) detectors based on the OMP algorithm are designed with the aid of a preset threshold. Specifically, different searching algorithms are designed, whose objective is to balance computational complexity and system performance. An upper bound for the average bit error probability (ABEP) of the first ECS detector is derived and used to optimize the preset threshold. Simulation results show that the proposed ECS detectors are capable of achieving a considerable reduction in computational complexity, compared with other near-optimal algorithms, with a negligible performance loss.