Fairness-Aware Inter-Slice Scheduler for IoT Services Over Satellite

This paper addresses the problem of scheduling the uplink bandwidth of Low Earth Orbit (LEO) satellites among multiple Internet of Things (IoT) slices with diverse Quality of Service (QoS) requirements. The scheduling process involves twofold decisions-the amount of bandwidth allocated and the allocation duration. Resource scheduling for satellite IoT services is challenging because of limited bandwidth availability during a satellite pass, especially for LEO satellites. Another challenge is to compute a fair allocation schedule for IoT services with different latency demands, packet transmission frequency, required data volume, and the number of IoT devices. To address these challenges, we propose a fairness-aware inter-slice scheduler for satellite IoT services in this work. The proposed scheduler computes service priority based on respective traffic demands. We propose two algorithms for the scheduler based on weighted greedy and Simulated Annealing (SA), respectively. The weighted greedy algorithm schedules the services greedily based on the priority order. The SA algorithm enhances the greedy solution by ensuring the allocated bandwidth is proportional to the respective priority values. The simulation results show that the proposed SA algorithm achieves up to 21.11% more proportional fairness than the Simulated Annealing and Monte Carlo (SA-MC) benchmark scheme.