Research Article
Dynamic Resource Allocation for Network Slicing in LEO Satellite Networks
@INPROCEEDINGS{10.1007/978-3-031-67162-3_8, author={Mingyu Zhu and Xiaofan Xu and Yueyue Zhang and Yihui Zhou and Ping Du and Du Xu and Xiaoning Zhang}, title={Dynamic Resource Allocation for Network Slicing in LEO Satellite Networks}, proceedings={Communications and Networking. 18th EAI International Conference, ChinaCom 2023, Sanya, China, November 18--19, 2023, Proceedings}, proceedings_a={CHINACOM}, year={2024}, month={8}, keywords={Satellite Networks Network Slicing Resource Allocation Software Defined Network Network Functions Virtualization}, doi={10.1007/978-3-031-67162-3_8} }- Mingyu Zhu
Xiaofan Xu
Yueyue Zhang
Yihui Zhou
Ping Du
Du Xu
Xiaoning Zhang
Year: 2024
Dynamic Resource Allocation for Network Slicing in LEO Satellite Networks
CHINACOM
Springer
DOI: 10.1007/978-3-031-67162-3_8
Abstract
Low earth orbit (LEO) satellite networks have advantages such as low latency and wide coverage, which play an increasingly significant role in global connectivity. Satellite network slicing is a technology that enables the partitioning of different resources and capabilities to different service requirements and customers in LEO satellite communication networks. It can guarantee different QoS requirements according to the application scenarios and traffic requirements. However, LEO satellite networks are characterized by uneven traffic distribution and certain periodicity over time, leading to large resource wastage on satellite networks. To address this problem, we study the resource allocation method in satellite network slices, which aims to optimize the utilization of idle communication resources and improve traffic deployment success rate in LEO satellite networks. In specific, we propose a heuristic algorithm named MMAS-RA based on max-min ant system algorithm for resource dynamic allocation. The experimental results in four typical scenarios illustrate that our method improves up to 32.26% traffic deployment success rate and 67.02% resource utilization rate compared with the benchmark algorithm.
