An Efficient User Clustering Framework for Non-Orthogonal Multiple Access Systems

Abstract

Introduction: This work introduces an efficient user clustering framework for Non-orthogonal multiple access (NOMA) systems  along with a structured, high-resolution Bluetooth-based dataset to facilitate advanced research in next generation wireless communication networks. As modern NOMA deployments increasingly rely on adaptive user grouping and interference-aware pairing, the lack of publicly available real world data has posed a significant limitation to empirical validation. In this paper, we focus on the capturing real-world signal characteristics in realistic indoor and urban environments, the dataset includes device-level metrics such as received signal strength indicator (RSSI), estimated distance, device type, and detection frequency parameters essential for modeling dynamic user heterogeneity and clustering behavior. Also, this paper presents a detailed classification of user clustering techniques specifically designed for NOMA-based wireless communication systems. Using these clustering schemes, we conducted extensive simulations to evaluate their performance in NOMA-based systems. The proposed clustering algorithm consistently outperforms existing techniques in terms of sum rate and energy efficiency (EE). The dataset serves as a critical asset for bridging theoretical models with field deployable NOMA architectures. It is particularly relevant for 5G and emerging 6G research, where intelligent access schemes must operate reliably under dense device populations and rapidly varying channels.