Abstract:In recent years, satellite communication systems have flourished. With the rapid increase of the number of users and the increase in QoS requirements, the single-layer satellite communication system has been far from meeting the needs. Heterogeneous satellite networks have become an inevitable development trend because they can achieve complementary advantages between constellations. The existing heterogeneous satellite network generally focus on heterogeneous networks between different orbit types, but the research on pure low-orbit satellite heterogeneous networks is rare. Therefore, this paper focuses on the research of low-orbit satellite heterogeneous networks. Among them, the routing algorithm is the basis of satellite interconnection, and its performance directly affects the performance of satellite network applications. In order to cope with common link congestion and satellite failure in heterogeneous satellite networks, this paper proposes a heterogeneous satellite network routing algorithm based on reinforcement learning and mobile agent. On the one hand, we use the reinforcement learning algorithm to adaptively reward the calculated optimal path, which is similar to ant pheromone accumulation. At the same time, the path causing congestion is appropriately punished. On the other hand, we use the inheritance of the mobile agent in orbit to solve the problem of inter-layer satellite failure. Theoretical analysis and simulation results show that reinforcement learning plays a key role in alleviating link congestion, and the greater the load, the better the effect. The mobile agent also effectively solves the problem of satellite failure. Compared with no mobile agent, the packet drop performance has been greatly improved.

Fully-Distributed Dynamic Packet Routing for LEO Satellite Networks: A GNN-Enhanced Multi-Agent Reinforcement Learning Approach

Inter-Satellite Routing for LEO Satellite Networks: A GNN and DRL Integrated Approach

GRouting: Dynamic Routing for LEO Satellite Networks with Graph-based Deep Reinforcement Learning

Graph Neural Network and Reinforcement Learning Based Routing for Mega LEO Satellite Constellations

Spatial Location Aided Fully-Distributed Dynamic Routing for Large-Scale LEO Satellite Networks

Distributed Routing Algorithm for LEO Satellite Network Based on Deep Reinforcement Learning

A GNN-Enabled Multipath Routing Algorithm for Spatial-Temporal Varying LEO Satellite Networks

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

Dynamic Routing Planning Method for Large-Scale Low-Orbit Satellite Networks Based on Location Guided and Multi-Agent DQN Network

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

Dynamic Routing for Integrated Satellite-Terrestrial Networks: A Constrained Multi-Agent Reinforcement Learning Approach

A DRL Empowered Multipath Cooperative Routing for Ultra-Dense LEO Satellite Networks

Enabling High-Throughput Routing for LEO Satellite Broadband Networks: A Flow-Centric Deep Reinforcement Learning Approach

Heterogeneous Satellite Network Routing Algorithm Based on Reinforcement Learning and Mobile Agent.

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

Distributed Routing Strategy Based on Machine Learning for LEO Satellite Network

Routing in Small Satellite Networks: A GNN-based Learning Approach

Fast-Convergence Reinforcement Learning for Routing in LEO Satellite Networks

Deep Reinforcement Learning-Based Multipath Routing for LEO Megaconstellation Networks

A Novel Routing Algorithm Combined with Multi-Agent System for LEO Satellite IP Networks

Knowledge Graph Aided Network Representation and Routing Algorithm for LEO Satellite Networks