Intelligent UAV and LEO-Assisted Edge Computing Systems for Real-Time IoT Applications

In this paper, we propose an intelligent edge computing system that supports unmanned aerial vehicles (UAVs) and low-Earth orbit (LEO) satellites for real-time utilization of Internet of Things (IoT) data within the space-air-ground integrated network (SAGIN) architecture. In this architecture, edge servers mounted on flying UAVs and LEO satellites provide the computing capability needed to process a large volume of collected IoT data. In the proposed system, our primary objective is to optimize the total energy consumed by the flying UAVs, considering their limited energy budget. To address this optimization challenge, we employ a joint optimization scheme that encompasses UAV trajectory and bit allocation, based on a successive convex approximation (SCA) algorithm. To assess the performance of our proposed approach, we compare the joint optimization scheme with various other optimization approaches in terms of total energy consumption.