Energy Balance-Oriented RF Energy Harvesting Schemes for Intelligent Reflecting Surface-Assisted Cognitive Radio Sensor Networks

Low and unbalanced energy harvesting (EH) levels among nodes severely restrict the network lifespan of cognitive radio sensor networks (CRSNs). To substantially extend network lifespan, this study introduces a subsurface partitioning-based intelligent reflecting surface (IRS) to assist the EH from a multiantenna sink to CRSNs nodes, and energy balance-oriented radio frequency (RF) EH schemes are proposed to increase fairness in EH across nodes by optimally configuring both the energy beamforming vector at the sink and the IRS passive beamforming vector. Specifically, with the IRS location being fixed, a constrained nonconvex optimization problem with the objective of maximizing the minimum EH level of all CRSNs nodes is formulated. An alternating optimization algorithm based on semidefinite relaxation and successive convex approximation is proposed to search for the joint solution. To boost the flexibility in configuring beamforming vectors, the EH duration is dynamically divided into variable-length time slots, within which both beamforming vectors can be flexibly adjusted. This strategy leads to a slight increase in the complexity of problem solving, but significantly equalizing the EH levels among nodes. Simulation results indicate that increasing the number of antennas at the sink, the number of IRS subsurface, collaboration among reflecting elements within the same subsurface, and dynamic division of EH duration all contribute to improving the performance of the EH schemes. Specifically, compared with the multiantenna benchmark schemes, the proposed RF EH schemes achieve at least a 40-fold increase in the minimum EH level and approximately an increase of 100% in terms of the EH fairness.