ABSTRACT
As 2-user non-orthogonal multiple access (NOMA) has been introduced in Chapter 5, we generalize the scenario to a multiple-user case. For a multi-user NOMA system, the joint optimization problem becomes more complicated because the high dimensional variables and power allocation variables have linear or convex constraints in general. As can be seen, the multi-user NOMA is different from the 2-user NOMA and brings new challenges.
For multi-user NOMA systems enabled by array beamforming, three different problems are considered in this chapter. In the second section, the user fairness for downlink NOMA networks is considered. To improve the overall data rate, the minimal achievable rate among multiple users is maximized. In the third section, joint Tx-Rx beamforming and power allocation in NOMA networks is considered. A boundary compresses particle swarm optimization (BC-PSO) algorithm is presented to solve this problem and a sub-optimal solution is obtained. In the fourth section, the hybrid analog-digital beamforming structure at the BS is considered, where the user grouping and the joint optimization of hybrid analog-digital beamforming and power allocation are explored to improve the sum rate of the users.
