Fast Resource Allocation for Downlink NOMA Based on Revenue and Chordal Graphs

Non-orthogonal multiple access (NOMA) is a promising technology for future mobile networks due to superior spectrum efficiency. In this paper, we model the interaction between the base station and multiple users as a Stackelberg game and devise a fast resource allocation method consisting of resource block allocation and power allocation. The objective is to serve as many users as possible at rates beyond their requirements, while seeking to enhance total revenue. We derive a closed-form formula for optimal power allocation. Besides, we convert the resource block allocation problem into the problem of finding a maximum weight independent set for a chordal graph, which only takes linear time. Simulation results show that the fast method we propose outperforms existing algorithms in total throughput and the number of users whose rate demands are attained.