Simultaneous Wireless Information and Power Transfer (SWIPT) in SAGIN

The Internet of Remote Things (IoRT) spans various critical applications such as smart agriculture, wildfire monitoring, and Internet of Medical Things, all demanding real-time data processing to maintain information freshness. However, executing these applications is highly challenging owing to the limited energy and computing resources of IoRT nodes. To support IoRT nodes in remote areas, integrated space-air-ground networks (SAGIN) can be efficiently designed. For instance, ground-based IoRT nodes can harvest energy and delegate computationally intensive tasks to low-altitude UAV swarms. These UAVs can also collaborate with High Altitude Platforms (HAPs) and Low Earth Orbit (LEO) satellites through task offloading. However, optimizing network performance in the dynamic and resource-constrained hierarchical environment of SAGIN remains complex. Additionally, owing to the dynamic time varying topology and constrained resources (i.e., energy, computing, caching, and bandwidth) in SAGIN, the network optimizations problem becomes very intricate to solve. In this project, our goal is to design light weight data driven adaptive machine learning algorithms that can jointly design trajectory for UAVs, UAV-IoRT device associations, timeslot allocation for energy harvesting and data transmission, offloading policy, and resource scheduling (i.e., Bandwidth, transmit power, computing and caching) to minimize age of information and energy consumption for both IoRTs and UAVs.