User Selection and Power Allocation for Detection-Performance Guarantee in Downlink Non-Orthogonal Multiple Access Systems

This paper investigates power allocation and user selection for a downlink non-orthogonal multiple access (NOMA) system with a base station and N users. A sum-capacity maximization problem for the single-input single-output (SISO) case is formulated in terms of power allocation among an arbitrary number of users under a total power constraint and a quality-of-service condition that the minimum rate requirement of each user is satisfied. A closed-form power allocation solution to the problem is derived and then used to develop a user selection method to maximize the number of users in a cluster for NOMA transmission subject to a detection performance requirement with guaranteed successive interference cancellation and a minimum transmit signal-to-noise ratio. Using full search for the optimization of user clustering among N users is complicated, so we sort all users according to their channel gains in descending order and determine the optimal number of NOMA users iteratively to reduce the complexity. At each iteration, all the users are partitioned into several groups and the leading user of each group is clustered for NOMA transmission. The combined user selection and power allocation method for the SISO case can be extended to a multiple-input multiple-output scenario. Computer simulation results are given to demonstrate the effectiveness of the proposed scheme.