Resource Scheduling for NOMA-Based Mobile Phones and IoT Fusion Satellite Network

With the advent of 5G, wireless communication has greatly changed people's lifestyles. However, limited by the coverage of terrestrial cellular networks, the demand for communication services of mobile phones in remote areas is difficult to be met. In order to meet the large coverage and quality of service (QoS) requirements of mobile phones, we consider a non-orthogonal multiple access (NOMA)-based mobile phones and internet of things (IoT) fusion satellite network. In this network, low earth orbit (LEO) satellite can significantly enlarge the coverage while NOMA and IoT further improve the utilization efficiency of resources. Aiming at maximizing the QoS satisfaction rate of mobile phones and the sum rate of internet of things devices, we formulate a joint optimization problem considering NOMA user grouping, inter-group power allocation and intra-group power allocation. To tackle the problem, we design a QoS priority joint optimization algorithm based on proximal policy optimization (Q-PPO). The experimental results show that the proposed Q-PPO algorithm always has considerable performance in QoS satisfaction rate and sum rate of internet of things devices.