Performance of Deep Embedded Clustering and Low-Complexity Greedy Algorithm for Power Allocation in NOMA-Based 5G Communication

This research probes a novel method of using Orthogonal Frequency Division Multiple Access based Non-Orthogonal Multiple Access (NOMA) to enhance downlink communication in fifth-generation (5G) networks. It explores the comparison between two sub-channel allocation methods namely the Low-Complexity Greedy Algorithm and Deep embedded clustering algorithm using measures like achievable rate and signal-to-noise ratios (SNR) calculated via Shannon's Capacity Theorem using Channel State Information (CSI), a dynamic power allocation technique. Simulation results verified through numerical evaluation show that proposed schemes can significantly increase the system capacity in a given scenario, creating new opportunities for advancements in wireless technology.