Dual Based Optimization Method for IRS-Aided UAV-Enabled SWIPT System

In this paper, we study a downlink unmanned aerial vehicle (UAV)-enabled simultaneous wireless information and power transfer (SWIPT) system with the aid of an intelligent reflecting surface (IRS). Considering the time switching (TS) scheme, the UAV is leveraged to charge a single battery limited smart device (SD) and for information transmission simultaneously. We aim to maximize the average harvested energy of the SD over a finite mission/communication period on the conditions of average data transmission requirement and other practical constraints, by jointly optimizing the UAV trajectory, TS factors, and the phase shifts of IRS. In existing literatures, the block coordinate descent and successive convex approximation (BCD-SCA) method is usually applied to obtain a local optimal solution, however, the solution is susceptible to the initial parameter settings. Therefore, in this paper, we obtain a more advantageous solution by using the Lagrange dual method. Numerical results show that under the same constraints, the proposed design scheme improves the average harvested energy compared with the BCD-SCA method, and complexity analysis indicates that our algorithm has lower computation complexity.