The Spectral Versus Energy Efficiency Trade-Off in Dynamic User Clustering Aided mmWave NOMA Networks

• Published in: IEEE transactions on communications Vol. 73; no. 6; pp. 4503 - 4519
• Main Authors: Rai, Sudhakar, Sharma, Ekant, Jagannatham, Aditya K., Hanzo, Lajos
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.06.2025; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Array signal processing; Clustering; Clustering algorithms; Energy efficiency; fractional programming; Heuristic algorithms; Hybrid power systems; Hybrid precoding; Interference cancellation; Maximization; Millimeter wave communication; Millimeter waves; MIMO; mmWave; Multiple objective analysis; Networks; NOMA; Nonorthogonal multiple access; Optimization; Resource management; Simulation; spectral efficiency; Tradeoffs; user clustering
• ISSN: 0090-6778; 1558-0857
• DOI: 10.1109/TCOMM.2024.3506920

The spectral efficiency (SE) and global energy efficiency (GEE) trade-off encountered in the design of millimeter-wave (mmWave)-based massive multi-input multi-output (MIMO) non-orthogonal multiple access (NOMA) networks is investigated with a particular focus on user clustering. By exploiting the similarity among user channels a pair of spectral and energy-efficient user clustering algorithms are proposed for dynamically selecting both the number of clusters and the number of users in each cluster. Subsequently, a joint analog precoder/combiner and user clustering technique is developed, followed by a multi-objective optimization (MOO) framework for flexibly balancing the GEE and SE objectives in a mmWave NOMA network subject to specific constraints. The MOO objective is initially transformed to a weighted sum rate maximization problem, followed by a quadratic-transform (QT)-based approach conceived for maximizing the non-convex objective by approximating it as a concave-convex function. Our simulation results demonstrate that the user clustering techniques designed attain a 85% performance gain over random clustering technique and demonstrating the benefits of the algorithm designed for mmWave NOMA networks.