Proportional Fairness-based Joint Channel and Power Allocation for Hybrid NOMA-OMA Downlink Systems

This work considers a downlink hybrid NOMAOMA multiuser transmission system that divides users into pairs, each pair sharing one channel using Non-Orthogonal Multiple Access (NOMA), while different pairs are assigned orthogonal channels using Orthogonal Multiple Access (OMA). To achieve high system efficiency while guaranteeing fairness, we propose a joint power and channel allocation framework with the proportional fairness objective, which maximizes the sum of logarithmic rates. The problem is decoupled into the power allocation (PA) and channel assignment (CA) subproblems, which are solved iteratively. Our main contribution is proposing a globally optimal solution algorithm for the CA subproblem, which is obtained by casting the problem as a bipartite graph matching problem. We show empirically that the proposed joint PA-CA solution performs very close to the exhaustive search for small numbers of users. Extensive experiments demonstrate that the proposed framework significantly outperforms several benchmark schemes in both system efficiency and fairness. Index Terms-NOMA, joint channel and power allocation, proportional fairness, sum of logarithmic rates, bipartite graph matching.