Power Allocation Based on Geometric Programming for Non-orthogonal Multiple Access Heterogeneous Networks

In practical situations, the performance of some power allocation schemes degrades sufficiently with imperfect channel state information (CSI). To account for the problem, we propose a novel power allocation scheme in Non-orthogonal multiple access (NOMA) heterogeneous networks based on imperfect CSI, which achieve optimal sum data rate while optimizing power allocation. It is proved that the optimization problem is non-convex with the constraint of outage probability limit. We first transform the probabilistic mixed problem into a non-probabilistic problem, which is solved based on geometric programming (GP), and then the power allocation for each small cell user is achieved via Lagrange dual method. The simulation results show that the proposed algorithm can significantly improve the output performance compared with equal-power allocation (EPA) algorithm.