On Multi-User Deep-Learning-Based Non-Coherent DPSK Multiple-Symbol Differential Detection in Massive MIMO Systems

Abstract

In view of reducing the complexity of signal detection in massive multiple-input multiple-output (MIMO) receivers, the use of non-coherent detection is favored over usual coherent techniques that require complex channel estimation. In this paper, a deep-learning approach to implement non-coherent multiple-symbol detection for differential phase-shift keying (DPSK) multi-user massive MIMO systems is proposed. The proposed deep-learning implementation reduces the high computational complexity required in conventional DPSK detection techniques. Two deep-learning-based multiple-symbol differential detection receiver designs are proposed and compared with decision-feedback differential detection (DFDD), and conventional multiple-symbol differential detection (MSDD) for the same system parameters. Multiple-symbol differential sphere detection (MSDSD) is used to implement conventional MSDD. The results show that the proposed deep-learning-based multiple-symbol differential detection implementation outperforms decision-feedback differential detection and achieves an optimal performance compared to conventional multiple-symbol differential detection implemented by multiple-symbol differential sphere detection for single-user and multi-user scenarios.

Keywords

• Deep-learning
• differential detection
• interference cancellation
• neural networks
• non-coherent detection
• massive MIMO