A blind detection method for spatial modulation in MIMO communications

Abstract

Spatial modulation (SM) is one of the probable candidates to be utilised in the fifth generation of wireless networks due to its power and spectral efficiencies. Since only one active transmit antenna exists in spatial modulation, inter‐channel interferences are avoided and the number of radio frequency (RF) chains is reduced. However, channel estimation is a major challenge in spatial modulation communication systems. In this study, a novel blind signal and channel estimation method for spatial modulation‐based multi‐input multi‐output communications, based on the well‐known Expectation‐Maximisation algorithm, is presented. This blind detector requires only a few pilot symbols at the beginning of the transmission to resolve the inherent phase and permutation ambiguities. Simulation results demonstrate that the proposed detector performs very similar to the optimum detector with full channel state information and outperforms existing blind detectors.

Blind detection for spatial modulation uplink multi‐user massive MIMO communications: A sparse Bayesian learning‐expectation maximization approach

“…Having determined such a three-dimensional matrix for a single-user scenario, each possible value of the matrix of latent variables corresponds to only a single component of the conditional pdf of P(Y|𝜃). 24 Inspired by what is defined in a single-user scenario, in multi-user uplink communications, there is a need for a three-dimensional matrix of latent variables with N t × M × N dimensions for each user. Consequently, the final matrix of latent variables for all users, called Z, would be a 2K + 1 dimensional matrix; where…”

“…Nonetheless, the parameters of this distribution should be estimated. These parameters are calculated in (24) and (25), utilizing the Bayes' theorem as 16…”

“…Assuming no CSI at the BS, a blind detector has been introduced in Reference 23, where the double sparsity of the transmitted signals and channel in massive GSM communications were considered. Based on the well‐known EM algorithm, a novel blind detector is proposed in Reference 24 for point‐to‐point SM communications. It has been shown in Reference 24 that the suggested blind detector performs similarly to the optimum Maximum Likelihood (ML) detector even in time‐varying fading channels.…”

“…Based on the well‐known EM algorithm, a novel blind detector is proposed in Reference 24 for point‐to‐point SM communications. It has been shown in Reference 24 that the suggested blind detector performs similarly to the optimum Maximum Likelihood (ML) detector even in time‐varying fading channels.…”

“…Inspiring by, 24 an uplink SM‐based multi‐user cellular network with time‐varying channel model is considered in this paper, where CSI is not available either at the BS or at the users. Thus, the present study is aimed to propose a blind detector for massive SM uplink communications in time varying channels, such that the detector in Reference 24 will be extended to the uplink scenario.…”