Priority-Aware General Packet Offloading in Multi-Layer Dense Satellite Networks

Multi-layer dense satellite networks (MDSNs) burgeons recently, and inter-satellite task offloading influenced by time-varying dynamic topology is an essential topic in MDSNs owing to appropriate scheduling scheme can significantly reduce the delay to improve the quality of service in MDSNs. However, the existing scheduling strategies primarily consider single-path transmission for each task in a time slot, which is unsuitable for large data amounts of tasks collected by earth observation satellites (OSs) in the system that pursues timeliness. In this paper, to effectively increase the QoS and utilize dense satellite resources, we propose the joint task-splitting and multiple path choosing (JTMPC) scheme related to transmitted links and combine OSs and communication satellites (CSs) resources to achieve tasks offloading. The constructed paths start with OSs, then go through multiple CSs simultaneously, and finally to ground stations for each task. Furthermore, we formulate the issue as a mixed integer nonlinear programming problem to minimize the end-to-end (E2E) delay by jointly considering path choosing, task-splitting ratio, queuing delay expenses, various topologies, and delay tolerance. Extensive analysis and numerical results corroborate that our proposed JTMPC scheme and the delay-oriented benders decomposition algorithm can achieve superior performance in E2E delay.