Block Deep Neural Network-Based Signal Detector for Generalized Spatial Modulation

Abstract

Generalized Spatial Modulation (GSM) is being considered for high capacity and energy-efficient networks of the future. However, signal detection due to inter channel interference among the active antennas is a challenge in GSM systems and is the focus of this letter. Specifically, we explore the feasibility of using deep neural networks (DNN) for signal detection in GSM. In particular, we propose a block DNN (B-DNN) based architecture, where the active antennas and their transmitted constellation symbols are detected by smaller sub-DNNs. After N-ordinary DNN detection, the Euclidean distance-based soft constellation algorithm is implemented. The proposed B-DNN detector achieves a BER performance that is superior to traditional block zero-forcing (B-ZF) and block minimum mean-squared error (B-MMSE) detection schemes and similar to that of classical maximum likelihood (ML) detector. Further, the proposed method requires less computation time and is more accurate than alternative conventional numerical methods.

Deep Learning-Aided 6G Wireless Networks: A Comprehensive Survey of Revolutionary PHY Architectures

“…CSI at the receiver is not necessary for fixed fading channels, but it is essential for time-varying channels. On the other hand, in [48], the authors propose a modeldriven DL-based block detector for GSM inspired by the linear block detectors such as Block-ZF and Block-MMSE detectors. The APM symbol of each active transmit antenna is detected by a DNN, assuming that perfect CSI is available at the receiver.…”

“…Studies discovered thus far unlock the potential of IM systems for future 6G and beyond communication technologies, owing to DL approaches that further enhance system efficiency by employing sophisticated algorithms such as TAS, power allocation, and MCS selection with remarkably low complexity. We may now conclude the examination of the literature on DL-aided MIMO systems and move on to the implementation of the generic MIMO-IM model presented in [48], which might give a sense of MIMO-DL programming to interested readers.…”

“…On the other hand, linear detectors, such as ZF and MMSE detectors, produce significantly worse BER performance than MLD, introducing a trade-off into GSM systems. To address this BER-complexity trade-off, the authors of [48] suggest a DNN-based detector for GSM systems. They consider a GSM system model with N t transmit and N r receive antennas (N r < N t ), where N p (2 ≤ N p N t ) out of N t transmit antennas are selected by index bits to be activated in any time slot.…”

“…However, complexity increases with the growing number of transmit antennas, as expected, and the block detection still performs worse than the optimal MLD due to the suboptimal linear part. Thus, the authors of [48] replace the linear component of the block detection with a DNN to enhance the accuracy of ŝI estimation. Fig.…”

DeepGOMIMO: Deep Learning-Aided Generalized Optical MIMO with CSI-Free Blind Detection

“…Finally, we evaluate the computational complexity of the proposed CSI-free blind DNN detector and compare it with other benchmark detectors, in terms of computation time [28]. For both the CSI-free blind DNN detector and the ZF-DNN detector, once the detector has been successfully trained, it can be used for MIMO detection for a long period of time without further retraining, unless the system parameters such as receiver location have been changed [22].…”