A two-dimensional OMA-NOMA user-pairing and power-minimization approach for opportunistic B5G-enabled IoT networks

The integration of cognitive radio (CR) into power domain non-orthogonal multiple access (NOMA) wireless systems has been envisioned as a promising paradigm for supporting massive energy-efficient beyond fifth-generation (B5G)-based IoT connectivity. This paper presents a novel two-dimensional orthogonal multiple access (OMA)-NOMA CR-based power minimization approach for battery-constrained B5G-based IoT networks. According to the proposed approach, the time and frequency domains are integrated together with NOMA for the purpose of serving a larger number of CR users. Our approach divides the available time of each idle frequency channel into a set of time slots, each can serve a group of CR users through power-domain multiplexing. Our approach consists of two phases: user-pairing with channel assignment and power-allocation. The user-pairing associated channel-assignment problem is mathematically presented as a binary linear-programming problem, which is shown to be a uni-modular optimization with polynomial-time complexity. Given the paired users and assigned channels, the power-minimization problem under a set of quality-of-service constraints is formulated as a non-convex optimization. Then, we implement an iterative method that transfers the problem into a convex optimization. Simulation results reveal that our two-dimensional OMA-NOMA CR-based approach provides significant performance gain over the conventional CR-based OMA-NOMA power-minimization in terms of the total consumed transmission power.