Latency Aware Resource Allocation and Task Offloading in a Hybrid GEO-LEO Satellite Network

Abstract

To accommodate the rapid development of the Satellite-Assisted Internet of Things (S-IoT) and the ever-shifting landscape of future communications, a hybrid satellite network integrating Geosynchronous Earth Orbit (GEO) and Low Earth Orbit (LEO) satellites has been proposed. This architecture is gaining prominence as a strong contender among emerging network architectures. The primary objective of this research is to optimize the network’s end-to-end energy consumption by addressing the latency of the network as a whole, cache capability, joint admission control, association, and power allocation. An equally pivotal facet involves ensuring fairness in IoT device association and impartial distribution of spectrum resources while simultaneously optimizing throughput. This distinctive objective, which remains relatively uncharted in prior research, encompasses the harmonization of fairness considerations alongside the pursuit of maximal efficiency. However, the hybrid GEO-LEO satellite network faces significant challenges due to limited onboard communication and computing resources, particularly in task offloading. The problem is classified as an NP-hard mixed-integer non-linear programming (MINLP) problem, which demands an effective solution approach with a low computational load. An outer approximation algorithm (OAA) is proposed to obtain a near-optimal solution to address this. The results demonstrate the ability of the approach to achieve fairness in IoT association, fairness in resource block (RB) allocation, and overall throughput in the hybrid GEO-LEO satellite network. This research contributes to advancing satellite networks, particularly the hybrid GEO-LEO architecture, to meet the future communication needs of the S-IoT. The proposed OAA algorithm provides a practical and efficient solution for the NP-hard MINLP problem, opening avenues for further advancements in this field.