Power Allocation in PDMA Systems with Imperfect Channel State Information

Mingyao Peng,Jie Zeng,Xin Su,Bei Liu
DOI: https://doi.org/10.1109/VTCFall.2019.8891563
2019-01-01
Abstract:Pattern division multiple access (PDMA) is a multi- carrier non-orthogonal multiple access (NOMA), which can meet the requirements of massive user connections and super-high data rate in the fifth generation (5G) wireless networks. In this paper, we work on the optimization of power allocation to improve the performance in the down-link PDMA system with imperfect channel state information (CSI) at transmitter. The outage throughput of the system is maximized by optimizing the power allocation under the constraints of maximum transmits power, minimum user data rate, and outage probability. Since this optimization problem is a probabilistic mixing problem, we first turn it into a non-probability problem. Then, assuming the pattern matrix is known, we propose an iterative power allocation scheme. The closed-form expression of power allocation is derived based on Karush-Kuhn-Tucker (KKT) conditions. The simulation results demonstrate that the proposed iterative power allocation scheme yields better performance over the existing schemes.
What problem does this paper attempt to address?