Deep learning based non-orthogonal multiple access system for multiple users

• Main Author: Tan, Khai Ming
• Other Authors: Teh Kah Chan
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Engineering; Deep learning; Non-orthogonal multiple access
• Online Access: https://hdl.handle.net/10356/176167
• Institution: Nanyang Technological University

The concept of Non-Orthogonal Multiple Access (NOMA) has been around for some time and since then, it has served as a viable solution to outperform the current conventional Orthogonal Multiple Access (OMA) whilst applying Deep Learning (DL) techniques. The report provides a smooth introduction into the world of wireless communications starting from its fundamentals, conventional NOMA, and DL-based NOMA. The main portion of this report explores the implementation and evaluation of a DL based approach for signal detection in an Orthogonal Frequency Division Multiplexing-Non-Orthogonal Multiple Access (OFDM-NOMA) system, specifically for a two-user uplink scenario. By leveraging the Long Short-Term Memory (LSTM) neural network architecture, we aim to enhance signal detection under various channel conditions, thereby addressing the challenges posed by conventional signal detection methods in NOMA systems. The Symbol Error Rate (SER) will be used as the metric across a range of signal-to-noise ratios (SNRs) to determine the overall system performance. The results reveal that the LSTM-based model exhibits superior performance in high noise environments, showcasing the potential of deep learning to improve the reliability and efficiency of signal detection in OFDM-NOMA systems. Four different configurations of batch sizes have also been applied to the training to determine the best performance with least trade-offs. This project not only demonstrates the capability of DL for complex signal classification tasks but also opens avenues for further research into the integration of machine learning algorithms in next generation wireless communication systems.