Intelligent Reflecting Surface-Assisted Wireless Powered Heterogeneous Networks

• Published in: IEEE transactions on wireless communications Vol. 22; no. 12; p. 1
• Main Authors: Zhu, Zhengyu, Li, Zheng, Chu, Zheng, Wu, Qingqing, Liang, Jing, Xiao, Yunlu, Liu, Peijia, Lee, Inkyu
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.12.2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Complexity; Convexity; Downlink; Downlinking; Energy consumption; Energy harvesting; Heterogeneous networks; Information transfer; intelligent reflecting surface (IRS); Internal energy; Internet of Things; Internet of Things (IoT); Kuhn-Tucker method; Mathematical analysis; Optimization; phase shifts; Reconfigurable intelligent surfaces; Riemann manifold; Scheduling; Semidefinite programming; Throughput; Uplink; Uplinking; Wireless communication; Wireless networks; Wireless powered heterogeneous networks (WPHN)
• DOI: 10.1109/TWC.2023.3274220
• ISSN: 1536-1276

In this paper, we introduce an intelligent reflecting surface (IRS)-assisted wireless powered heterogeneous network (WPHN) consisting of two heterogeneous groups of devices. Specifically, one group of devices, i.e., energy-harvesting devices (EHDs), are charged by external energy supplies, while the other group of devices, i.e., non-energy-harvesting devices (NEHDs), are powered by internal energy supplies. An IRS aims to participate in the wireless energy transfer (WET) in downlink and the wireless information transfer (WIT) in the uplink. A sum throughput maximization problem is formulated subject to the constraints of individual energy consumption, transmission time scheduling, and IRS phase shifts. To cope with the non-convexity of the problem, we first derive the optimal IRS phase shifts of the uplink WIT independently. Next, the semi-definite programming (SDP) relaxation is adopted to recast this non-convex problem into the convex one, which can be numerically solved. Then, a novel low-complexity scheme is developed to gain more insights and mitigate the computational complexity induced by the SDP relaxation. In particular, the dual problem and Karush-Kuhn-Tucker conditions are first utilized to obtain the optimal transmission time scheduling. Then, we propose a method based on Riemannian manifold optimization to compute the optimal IRS phase shifts of the downlink WET in closed-form. Finally, simulation results are presented to verify the optimality of our proposed scheme, and highlight the benefits induced by the IRS to coordinate these heterogeneous devices.