Energy Efficiency Optimization for PSOAM Mode-Groups Based MIMO-NOMA Systems

Plane spiral orbital angular momentum (PSOAM) mode-groups (MGs) and multiple-input multiple-output non-orthogonal multiple access (MIMO-NOMA) serve as two emerging techniques for achieving high spectral efficiency (SE) in the next-generation networks. In this paper, a PSOAM MGs based multi-user MIMO-NOMA system is studied, where base station transmits data to users by utilizing generated beams. For such scenario, interference between different mode groups can be avoided, which leads significant performance enhancement. We aim maximize energy (EE) of subject constraints total transmission power minimum rate. This designed optimization problem non-convex owing among users, hence quite difficult tackle directly. To solve issue, we develop dual layer resource allocation algorithm bisection method exploited outer obtain optimal EE distributed iterative inner optimize transmit power. Besides, an alternative with Deep Belief Networks (DBN) proposed cope requirement low computational complexity. Simulation results verify theoretical findings demonstrate algorithms on better comparing conventional terms EE.