Throughput Maximization for IRS-aided UAV-powered Green IoT Network

As an essential technology for constructing green passive Internet of Things (IoT) network, backscatter communication (BackCom) enables battery-free IoT devices to deliver information by modulating and reflecting incident carriers. Nevertheless, the energy harvesting at an IoT device is constrained by its distance from the radio-frequency (RF) emitter, and the double-fading phenomenon significantly restricts the achievable performance of BackCom-based IoT network. Existing research has revealed that intelligent reflecting surface (IRS) and dynamic unmanned aerial vehicle (UAV) are promising solutions to overcome the current bottleneck, and their combined application in IoT network for BackCom remains in the preliminary exploration phase. Most studies focus more on traditional fixed RF emitters, which lack the flexibility needed for efficient energy transmission and are not suitable for infrastructure blank areas or hard-to-maintain regions. Motivated by this, an IRS-aided UAV-powered green IoT network is proposed in this paper, where the UAV acts as a mobile power beacon to energize multiple IoT nodes on the ground in a time-division multiple address manner, and an IRS is deployed in the scenario to enhance the BackCom performance. To guarantee reliable data transfer under energy harvesting constraint, we maximize the minimum average throughput among all IoT nodes for BackCom during the UAV flight duration by optimizing the node communication scheduling, power splitting coefficient, UAV trajectory and IRS phase shift. Specifically, we employ a four-stage alternating optimization method to decouple the optimization variables and solve for each variable independently. Finally, extensive experimental results verify the superior performance of our proposal in improving throughput over other benchmark schemes.