Optimizing resource allocation in 5G wireless networks for enhanced spectral efficiency and energy conservation using machine learning methods

Due to rising data needs and technological developments, the wireless communication industry undergoes a revolution every 10 years. The main goal of the 4G–5G transition is to increase gigabit capacity while reducing power consumption. A foundation in communication theory is necessary to achieve the big goals of 5G, which include huge capacity spikes, efficient spectrum usage, and low power use. Faster data transmissions are needed for future wireless technology. Two emerging potential technologies are non-orthogonal multiple access (NOMA) and massive multi-input and multi-output (MIMO). Resource distribution at the base station (BS) must be done efficiently to provide equitable service to customers. Increased capacity is required by wireless networks to handle an increasing user base. Integration of MIMO-NOMA is one important approach. The distribution of electricity in particular is a research hotspot that is influencing 5G success. This novel method simplifies downlink NOMA power distribution by combining the salp swarm algorithm (SSA) and crowd search algorithm (CSA). This improves spectrum utilization, power efficiency, and throughput. The hybrid algorithm delivers outstanding energy efficiency (EE) results at signal-to-noise ratios (SNRs) of SEs of 3.274bit/s/Hz, 7.282bit/s/Hz, and 11.84bit/s/Hz for M = 2,20, and 150. Significant power improvements are highlighted by the max EE values, which reach 1.651,10–4,135 bit/Joule, 1.949,10–4,6497bit/Joule.