Priority-Aware Parallel Transmission Toward Dense Satellite Remote Sensing and Communication Integrated Networks

Dense satellite networks provide new potentials for prompt massive observational data backhaul, which has been the focus of the study. However, the dynamic and dense networks, coupled with the multi-priority task requirements of satellites, present significant challenges in designing effective offloading and transmission strategies. To address these challenges, we construct a remote sensing and communication integrated network (RSCIN) model and propose a task-splitting and parallel transmission approach that adequately utilizes the resources of both communication satellite (CS) and observation satellite (OS) for efficient data offloading. Specifically, we first investigate the priority-aware latency caused by the preemptive-resume scheme of OSs and employ a lognormal distribution to model the internal traffic intensity of CSs and analyze its influence on OS data relays. Furthermore, we formulate a mixed integer nonlinear programming (MINLP) problem to minimize the end-to-end (E2E) delay by jointly considering path selection, task-splitting strategy, transmit power, and queuing delay. With the proposed joint task-splitting and multi-path selection (JTMPS) algorithm, we equivalently decompose the MINLP problem into the constructed path set (CPS) problem and an optimal CPS-based task scheduling problem, which the benders decomposition algorithm can further solve. Extensive analysis and numerical results verify that the proposed JTMPS algorithm can achieve superior performance than various baseline schemes in RSCINs.