Power Optimization in RIS-Assisted SWIPT-IoT System With Discrete Phase Shift

The integration of reconfigurable intelligent surfaces (RIS) and simultaneous wireless information and power transfer (SWIPT) present a promising solution for sustainable and efficient wireless communications in large-scale IoT networks, especially within energy-constrained smart agriculture applications. However, most existing works assume ideal energy harvesting (EH) models and continuous RIS phase shifts that limit their practical relevance. This paper addresses these limitations by proposing a total transmit power minimization framework for a multi-user RIS-assisted MISO power-splitting (PS) SWIPT system. First, a practical logistic non-linear energy harvesting (NL-EH) model is adopted to better reflect the realistic behavior of RF energy conversion circuits. Second, a discrete phase shift (DPS) model with finite quantization levels is employed to account for practical RIS hardware constraints. Third, an alternating optimization algorithm is developed for the resulting non-convex optimization problem through joint optimization of the base station’s beamforming vectors, RIS reflection matrix, and PS ratio. Techniques such as Zero-Forcing (ZF), Semidefinite Relaxation (SDR), and Gaussian randomization (GR) are leveraged to address the associated sub-problems, while an alternate one-dimensional search strategy is used for RIS phase shift optimization. Finally, numerical simulations are conducted to validate the proposed framework in terms of performance and convergence. The results demonstrate robustness under imperfect channel state information (ICSI) and varying system parameters for next-generation IoT-enabled smart farming use-case.