Abstract
Massive spatial modulation (MSM) is considered as an attractive technique for multi antenna wireless communications. That is because, it gives higher energy efficiency and spectral efficiency than small scale multiple-input multiple-output (MIMO) systems. Massive SM-MIMO utilizes multiple transmit antennas for each user with only one transmit radio frequency (RF) chain and hundreds of receive antennas at base station (BS) with small number of RF chain. Where, each user can activate any one of its transmit antennas and the index of active transmit antenna (TA) can convey to information bits in addition to the information bits conveyed through classical modulation symbols (e.g. 16QAM). Owing to large number of TAs at the user and small number of RF chains at BS, multi user signal detection becomes challenging problem. To solve this matter, a joint grouped SM transmission scheme at users and group subspace pursuit (GSP) based signal detection at BS can be proposed to improve the signal detection performance. Owing to joint transmission scheme, SM signals in same transmission group exhibit group sparsity. Also, spatial signal composed of multiple users' SM signals exhibits distributed sparsity. By utilizing these sparse features, the proposed GSP based signal detection can detect SM signals more reliability than other detection techniques. Additionally, the cyclic prefix single carrier (CPSC) is utilized to withstand the multipath channels. Simulation results prove that BER performance of the proposed GSP based signal detection outperforms classical SP based signal detection by 8 dB SNR gain at BER = 10(-2). This gain can be improved by 2 dB by increasing the number of transmit antenna.