Congestion Avoidance Geographic Routing in a Large-Scale Multiple Shell Low Earth Orbit Satellite Constellation

With the research and prospects of 6G technology, large-scale low-earth orbit satellite networks will play a crucial role in future communication frameworks. In this study, we proposed the congestion avoidance geographic routing algorithm, utilizing an online algorithm to dynamically adapt to changes in network topology and solve the network congestion problem. The algorithm not only determines the forwarding satellites based on the distribution of neighboring satellites but also employs traffic thresholds to avoid heavily loaded satellites. Simulation results indicate that our algorithm effectively avoids congestion and achieves load balancing throughout the network. Compared to traditional geographic routing, our method achieves lower end-to-end delay.

Introduction

In recent years, deploying low-earth orbit (LEO) satellite networks like Starlink, OneWeb, and Kuiper has provided global, ubiquitous broadband services. Compared to traditional geostationary orbit (GEO) satellites, LEO satellites offer advantages such as shorter transmission delays, lower link power, smaller size, and reduced launch costs. The lower launch costs of LEO satellites allow for the deployment of many satellites, achieving high coverage areas. However, due to their small size, LEO satellites have limitations such as reduced battery capacity, lower payload capabilities, and inferior computational power. Additionally, satellites in the ascending subshell and descending subshell cannot communicate with each other due to their different moving directions and high relative speeds, posing various challenges for routing on LEO satellite networks.