Hurry: Dynamic Collaborative Framework For Low-orbit Mega-Constellation Data Downloading

• DOI: 10.48550/arxiv.2406.03159
• Publication Date: 2024-06-05
• Authors: Handong Luo; Wenhao Liu; Qi Zhang; Ziheng Yang; Quanwei Lin; Wenjun Zhu; Kun Qiu; Zhe Chen; Yue Gao
• Keywords: Mega-; Upload; Orbit (dynamics)

Low-orbit mega-constellation network, which utilize thousands of satellites to provide a variety network services and collect wide range space information, is rapidly growing field. Each satellite collects TB-level data daily, including delay-sensitive used for crucial tasks, such as military surveillance, natural disaster monitoring, weather forecasting. According NASA's statement, these need be downloaded the ground processing within 3 5 hours. To reduce time required downloads, state-of-the-art solution known CoDld, only available small constellations, uses an iterative method cooperative downloads via inter-satellite links. However, in LMCN, download same amount using CoDld will exponentially increase compared downloading constellation. We have identified analyzed reasons this degradation phenomenon propose new framework, named Hurry. By modeling mapping topology changes transmission Time-Expanded Graphs, we implement our algorithm Hurry framework avoid effects. In fixed volume evaluation, achieves 100% completion task while reached 44% progress. continuous generation flow improves throughput from 11% 66% different scenarios.