Abstract:Low-Earth orbit (LEO) satellites have limited on-board resources, user terminals are unevenly distributed in the constantly changing coverage area, and the service requirements vary significantly. It is urgent to optimize resource allocation under the constraint of limited satellite spectrum resources and ensure the fairness of service admission control. Therefore, we propose an intelligent hierarchical admission control (IHAC) strategy based on deep reinforcement learning (DRL). This strategy combines the deep deterministic policy gradient (DDPG) and the deep Q network (DQN) intelligent algorithm to construct upper and lower hierarchical resource allocation and admission control frameworks. The upper controller considers the state features of each ground zone and satellite resources from a global perspective, and determines the beam resource allocation ratio of each ground zone. The lower controller formulates the admission control policy based on the decision of the upper controller and the detailed information of the users' services. At the same time, a designed reward and punishment mechanism is used to optimize the decisions of the upper and lower controllers. The fairness of users' services admissions in each ground zone is achieved as far as possible while ensuring the reasonable allocation of beam resources among zones. Finally, online decision-making and offline learning were combined, so that the controller could make full use of a large number of historical data to learn and generate intelligent strategies with stronger adaptive ability while interacting with the network environment in real time. A large number of simulation results show that IHAC has better performance in terms of a successful service admission rate, service drop rate, and fair resource allocation. Among them, the number of accepted services increased by 20.36% on average, the packet loss rate decreased by 17.56% on average, and the resource fairness increased by 17.16% on average.

Reinforcement Learning Based Control Domain Division in LEO Satellite Networks

Dynamic Controller Placement for Software-Defined LEO Network Using Deep Reinforcement Learning

Dynamical Control Domain Division for Software-Defined Satellite-Ground Integrated Vehicular Networks.

Reinforcement Learning Based Intelligent Routing for Software Defined LEO Satellite Networks

Multi-Commodity Flow Routing for Large-Scale LEO Satellite Networks Using Deep Reinforcement Learning

Reinforcement learning based dynamic distributed routing scheme for mega LEO satellite networks

Collaborative Computing in Non-Terrestrial Networks: A Multi-Time-Scale Deep Reinforcement Learning Approach

Mobility- and Load-Adaptive Controller Placement and Assignment in LEO Satellite Networks.

Reinforcement Learning for Cognitive Delay/Disruption Tolerant Network Node Management in an LEO-based Satellite Constellation

Adaptive Network Management Service Based on Control Relation Graph for Software-Defined LEO Satellite Networks in 6G

Deep Reinforcement Learning for Congestion Control and Routing Optimization in LEO Satellite Networks

A Dynamic Migration Strategy of SDN Controllers in LEO Networks

Collaborative Deep Reinforcement Learning for Resource Optimization in Non-Terrestrial Networks

An Intelligent Routing Algorithm for LEO Satellites Based on Deep Reinforcement Learning

Deep Reinforcement Learning for Wireless Scheduling in Distributed Networked Control

Software Defined Routing Algorithm in LEO Satellite Networks

Load-Aware Dynamic Controller Placement Based on Deep Reinforcement Learning in SDN-Enabled Mobile Cloud-Edge Computing Networks

SDN Controller Load Balancing Based on Reinforcement Learning

Optimized Controller Provisioning in Software-Defined LEO Satellite Networks

Intelligent Hierarchical Admission Control for Low-Earth Orbit Satellites Based on Deep Reinforcement Learning