DEEP LEARNING BASED SIGNAL DETECTION AND CHANNEL ESTIMATION FOR MIMO-NOMA SYSTEM

Abstract

The increasing demands for enormous connectivity, low latency, and high reliability of future communication networks require new techniques. Multiple-input-multiple-output and non-orthogonal multiple access (MIMO-NOMA), which incorporate the NOMA concept into MIMO, is an appealing technology to reinforce system capacity and spectral efficiency in future communication scenarios. However, rapidly changing channel conditions and high computational complexity due to SIC degrade system performance. To tackle these limitations, this paper proposes a deep learning-based MIMO-NOMA framework for data detection and estimation of sharply changing channel conditions. Specifically, an effective deep neural network is designed for communication, incorporating several convolutional layers and multiple hidden layers. The NOMA-MIMO-DL framework addresses the data detection problem to achieve a lower signal error rate and automatically detect the channel characteristics of MIMO-NOMA. In general, the proposed cooperative framework is built, trained, and tested to enable automatic encoding, decoding, and channel detection in a relay feeding channel. Simulation results demonstrate that the proposed scheme is robust and efficient compared to standard approaches.