RIS-Assisted SWIPT for IoT Networks: A Survey of Architectures, Optimization, and Future Directions

Abstract

The fusion of Simultaneous Wireless Information and Power Transfer (SWIPT) with Reconfigurable Intelligent Surfaces (RISs), i.e., RIS-aided SWIPT, is emerging as a promising paradigm for sustainable, energy-efficient, and intelligent sixth-generation (6G) Internet of Things (IoT) networks. By jointly enabling programmable propagation, Information Decoding (ID), and radio-frequency Energy Harvesting (EH), RIS-assisted SWIPT can extend the lifetime of energy-constrained IoT devices while improving coverage, reliability, and resource utilization. Hence, this paper presents a unified IoT-oriented survey of state-of-the-art RIS-assisted SWIPT systems, categorizing research advances according to system models, SWIPT receiver architectures, RIS deployment architectures, optimization methodologies, practical impairments, and IoT application requirements. We review various RIS implementations, including Passive RIS (PRIS), Active RIS (ARIS), Simultaneous Transmitting and Reflecting RIS (STAR-RIS), Beyond-Diagonal RIS (BD-RIS), Extremely Large RIS (XL-RIS), self-sustainable RIS, multi-RIS, and hybrid RIS deployments. We also discuss emerging 6G technologies, including movable antenna-assisted SWIPT, pinching antenna-assisted SWIPT, movable RIS, fluid antenna-assisted RIS systems, holographic Multiple Input Multiple Output (MIMO) surfaces, and RIS-enabled semantic communications. Special attention is devoted to the evolution of optimization and learning-based approaches, including Semidefinite Relaxation (SDR), Successive Convex Approximation (SCA), Fractional Programming (FP), Alternating Optimization (AO), Block Coordinate Descent (BCD), Majorization-Minimization (MM), Inner Approximation (IA), Riemannian Manifold Optimization (RMO), robust optimization, Deep Reinforcement Learning (DRL), Graph Neural Networks (GNNs), Federated Learning (FL), and meta-learning. The survey further connects these algorithms to the core mathematical structure of RIS-assisted SWIPT problems, including coupled active/passive beamforming, nonlinear energy-harvesting (NL-EH) constraints, imperfect Channel State Information (CSI), discrete phase shifts, hardware impairments, RIS power consumption, synchronization, scalability, and deployment cost. In addition, we provide an IoT-oriented discussion of application-specific requirements in smart healthcare, industrial automation, intelligent transportation, smart agriculture, and environmental monitoring. Through taxonomy, comparative synthesis, lessons learned, and cross-paper analysis, this survey highlights the trade-offs among Energy Efficiency (EE), Spectral Efficiency (SE), harvested energy, robustness, complexity, and practical deployability. Finally, we present a forward-looking research roadmap highlighting future directions for RIS-SWIPT systems, including AI-driven resource allocation, integration with Integrated Sensing and Communication (ISAC), security-aware transmission, Rate Splitting Multiple Access (RSMA), and Uncrewed Aerial Vehicles (UAV)-RIS architectures to enable scalable and sustainable massive IoT deployments in 6G networks.