User Association in Non-Orthogonal Multiple Access Networks

In this paper, the flexible user association is investigated in non-orthogonal multiple access (NOMA)-based multiple base stations (BSs) networks. More particularly, users are partitioned into multiple orthogonal clusters to allocate into different resource blocks (RBs) for avoiding serious co-channel interferences. In an effort to maximize the weighted sum rate of the system, a user association optimization problem is formulated. Two algorithms based on coalitional game are proposed for obtaining suboptimal and global optimal solutions, respectively. The properties of the proposed algorithms, including complexity, convergence, stability and optimality are analyzed. Simulation results demonstrate that: i) the proposed algorithms can significantly enhance the weighted sum rate of the considered systems; ii) the proposed NOMA-based system is capable of achieving promising gains over conventional orthogonal multiple access (OMA)-based framework, and iii) the considered distance based weight factor can greatly improve the fairness of users.