Joint transmission design for IRS-assisted MISO SWIPT systems
Zhen Liu,Xiaodong Zhu,Baojun Chen,Xiaodong Tu,Jun Xie,Zhonglou Meng
DOI: https://doi.org/10.1016/j.sigpro.2022.108649
IF: 4.729
2022-11-01
Signal Processing
Abstract:In this paper, we investigate the transmission design for an intelligent reflecting surface (IRS)-assisted multiple-input single-output (MISO) simultaneous wireless information and power transfer (SWIPT) system, where an IRS adjusts its reflection phase shifts to facilitate information transfer and energy harvesting. Two types of input signals, i.e., Gaussian signals and finite-alphabet signals, are both considered. Our goal is to maximize the mutual information between the access point (AP) and the information receiver (IR) by jointly optimizing the transmit beamforming at the AP and the phase shifts of the IRS under the constraints of transmit power at the AP and required harvested energy at the energy harvesting receiver (ER). We show that the transmission designs for these two types of signals actually can be unified into a common optimization problem. Although the formulated problem is an intractable non-convex problem, the solving frameworks in cases of perfect and imperfect channel state information (CSI) are provided, respectively. For the perfect case, an efficient algorithm, which combines semidefinite relaxation (SDR) and penalty-based manifold optimization (PMO), is proposed. For the imperfect case, we study the worst-case transmission design and develop an algorithm based on SDR. Numerical results show the superiority of our proposed algorithms.
engineering, electrical & electronic