QoE-Centric Resource Management for NOMA-Based LEO Satellite–Terrestrial Integrated Networks

In this article, we investigate the resource management for Quality of Experience (QoE)-centric low Earth orbit (LEO) satellite–terrestrial integrated network (STIN) based on nonorthogonal multiple access (NOMA). In satellite communication, existing efforts are devoted to maximizing Quality of Service (QoS) metrics, such as throughput. The optimization mechanism centered around QoS may to some extent impact the fairness of the system, leading to a decrease in resource utilization. To achieve more rational resource allocation, we focus on maximizing the user's QoE. We apply NOMA to LEO STIN to improve spectral efficiency and utilize the mean opinion score (MOS) to assess user satisfaction. Then, we maximize the sum MOSs of all users by jointly optimizing channel access, user pairing, and power allocation. However, the proposed problem is a complex nonconvex problem that is challenging to solve directly. To tackle this issue, we propose a coalition formation algorithm based on better response for user pairing and channel access. The power allocation is performed by deriving and presenting a linear search algorithm. Simulation results show that the proposed QoE optimization framework can effectively improve the QoE performance of the system compared with the QoS-centric optimization scheme and the OMA transmission method. Furthermore, in contrast to existing work, the QoE optimization proposed in this article enhances the fairness of the system and increases resource utilization without incurring additional system costs.