Joint user clustering and salp based particle swarm optimization algorithm for power allocation in MIMO-NOMA

• Published in: Journal of intelligent & fuzzy systems Vol. 40; no. 5; pp. 9007 - 9019
• Main Authors: Subudhi, Jyotirmayee, Indumathi, P.
• Format: Journal Article
• Language: English
• Published: London, England SAGE Publications 01.01.2021; Sage Publications Ltd
• Subjects: Algorithms; Clustering; Convergence; Energy conversion efficiency; Energy efficiency; Evolutionary algorithms; Nonorthogonal multiple access; Optimization; Particle swarm optimization; Quality of service architectures; PSO optimization; user clustering; power allocation; Energy efficiency; layered transmission; MIMO-NOMA; Non-orthogonal multiple access
• ISSN: 1064-1246; 1875-8967
• DOI: 10.3233/JIFS-201412

Non-Orthogonal Multiple Access (NOMA) provides a positive solution for multiple access issues and meets the criteria of fifth-generation (5G) networks by improving service quality that includes vast convergence and energy efficiency. The problem is formulated for maximizing the sum rate of MIMO-NOMA by assigning power to multiple layers of users. In order to overcome these problems, two distinct evolutionary algorithms are applied. In particular, the recently implemented Salp Swarm Algorithm (SSA) and the prominent Optimization of Particle Swarm (PSO) are utilized in this process. The MIMO-NOMA model optimizes the power allocation by layered transmission using the proposed Joint User Clustering and Salp Particle Swarm Optimization (PPSO) power allocation algorithm. Also, the closed-form expression is extracted from the current Channel State Information (CSI) on the transmitter side for the achievable sum rate. The efficiency of the proposed optimal power allocation algorithm is evaluated by the spectral efficiency, achievable rate, and energy efficiency of 120.8134bits/s/Hz, 98Mbps, and 22.35bits/Joule/Hz respectively. Numerical results have shown that the proposed PSO algorithm has improved performance than the state of art techniques in optimization. The outcomes on the numeric values indicate that the proposed PSO algorithm is capable of accurately improving the initial random solutions and converging to the optimum.