Optimal Water-Filling Algorithm in Downlink Multi-Cluster NOMA Systems

TL;DR

This paper introduces a globally optimal power allocation algorithm for multi-cluster NOMA systems, leveraging a water-filling approach to maximize sum-rate and improve efficiency over existing methods.

Contribution

It proposes a novel joint intra- and inter-cluster power allocation method using a closed-form solution and a water-filling algorithm, enhancing NOMA system performance.

Findings

The algorithm achieves higher sum-rate compared to traditional NOMA and OMA.

Each NOMA cluster can be treated as a virtual OMA user with a specific effective channel gain.

Simulation shows significant performance improvements in sum-rate and outage probability.

Abstract

The key idea of power-domain non-orthogonal multiple access (NOMA) is to exploit the superposition coding (SC) combined with successive interference cancellation (SIC) technique (called SC-SIC) while reducing the receivers' complexity as well as error propagation. Actually, NOMA suggests a low-complexity scheme, where users are grouped into multiple clusters operating in isolated resource blocks, and SC-SIC is performed among users within each cluster. In this paper, we propose a globally optimal joint intra- and inter-cluster power allocation for any arbitrary user grouping to maximize users' sum-rate. In this algorithm, we exploit the closed-form of optimal intra-cluster power allocation obtained in our previous work. Then, by transforming network-NOMA to an equivalent virtual network-OMA, we show that the optimal power allocation can be obtained based on the very fast water-filling algorithm. Interestingly, we observe that each NOMA cluster acts as a virtual OMA user whose effective channel gain is obtained in closed form. Also, each virtual OMA user requires a minimum power to satisfy the minimum rate demand of its real multiplexed user. In simulation results, we evaluate the performance gap between fully SC-SIC, NOMA, and OMA, in terms of users sum-rate, and outage probability.