Low Earth Orbit (LEO) satellite networks are favored for their global coverage, low latency, and geographical flexibility, but they also face challenges of high dynamism in network topology and congestion control. To address these issues, this study introduces an improved BBR congestion control algorithm (Deep-BBR) based on traffic prediction. The algorithm integrates Long Short-Term Memory networks, Graph Convolutional Networks, and attention mechanisms to accurately predict bandwidth utilization and dynamically adjust pacing gains. Experimental results show that the Deep-BBR algorithm significantly outperforms the traditional BBR algorithm in effective throughput, latency, and in-flight data management, effectively enhancing the communication performance and stability of LEO satellite networks. This improvement offers a more efficient and intelligent solution for congestion control in LEO satellite networks.
An Improved BBR Algorithm with Adaptive Congestion Control for LEO Satellite Networks
31.05.2024
1259597 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Multimedia traffic management and congestion control in satellite ATM networks
DSpace@MIT | 1994
|Algorithm improved congestion propagation analysis method
Europäisches Patentamt | 2022
|