Abstract
As wireless networks evolve toward 5G and 6G, the integration of Simultaneous Wireless Information and Power Transfer (SWIPT) becomes critical for supporting energy-constrained IoT devices. However, SWIPT faces significant challenges regarding physical layer security and hardware complexity. This paper proposes a robust transmit time-switching (TS) protocol that jointly optimizes beamforming and timing to provide secure data transmission and efficient energy harvesting. By utilizing energy signals as artificial noise to jam eavesdroppers, the proposed approach achieves superior secrecy energy efficiency compared to traditional power-splitting methods, even under imperfect channel state information (CSI).
Introduction
The proliferation of Internet of Things (IoT) devices has spurred interest in SWIPT, where a base station (BS) concurrently transmits data and power. Despite its potential, SWIPT is hindered by two primary issues: security vulnerabilities due to the broadcast nature of wireless signals and the high hardware complexity of receiver-side power-splitting (PS) architectures. Furthermore, practical systems must contend with imperfect CSI, which often degrades the performance of standard beamforming algorithms.
This research addresses these hurdles by introducing a Transmit Time-Switching (TS) framework. Unlike PS, which requires complex hardware to divide signals at the receiver, TS alternates between energy and information phases at the transmitter. This design not only simplifies hardware but also enhances security by using energy-carrying signals to jam potential eavesdroppers.
System Model and Methodology
The proposed system considers a multicell network where each BS serves multiple legitimate users while being monitored by potential eavesdroppers.
Multi-Purpose Beamforming
The BS alternates between an "energy" phase and an "information" phase. During the energy phase, beamformers are designed not just to maximize power delivery to legitimate users, but to simultaneously act as a jamming signal for eavesdroppers. This "energy-as-noise" strategy leverages the inherent overhead of power transfer to enhance physical layer security.
Joint Optimization and Robust Design
The researchers developed a mathematical framework to jointly optimize the energy beamforming vector, the information beamforming vector, and the time-switching ratio ($\eta$). To handle the non-convex nature of the optimization problem, a novel path-following algorithm was designed. This algorithm decomposes the problem into a sequence of Convex Quadratic Programs (CQP), ensuring fast convergence and low computational overhead. To account for real-world inaccuracies, "worst-case" uncertainty models were integrated into the robust design, maintaining performance despite channel estimation errors.
Performance Evaluation
The proposed algorithms were evaluated using Monte Carlo simulations of a three-cell network. Each cell contained four users (two near the BS and two far) and one multi-antenna eavesdropper. The simulations utilized Rician fading channels with a path loss exponent of 3, noise variance at -90 dBm, and BS transmit power ranging from 26 to 30 dBm.
Key Results
- Secrecy Rate Gains: The TS-based algorithm outperformed leading PS-based methods, achieving a secrecy rate gain of approximately 0.5 bits/sec/Hz.
- Efficiency and Scalability: The optimization algorithm typically converges within 20–25 iterations. Furthermore, the secrecy rate scales almost linearly with the number of BS antennas, indicating suitability for Massive MIMO.
- Robustness: The system demonstrated remarkable stability against CSI uncertainty. Even as estimation errors increased, the drop in secrecy rate remained minimal compared to non-robust alternatives.
Conclusion and Future Work
This work demonstrates that Transmit Time-Switching is a viable and secure alternative to Power Splitting for SWIPT systems. By repurposing energy signals for jamming, the proposed method achieves high secrecy energy efficiency with reduced hardware complexity.
Future research will focus on integrating non-linear energy harvesting models, as current EH circuits exhibit saturation points not captured by linear models. Additionally, the team aims to investigate scenarios where eavesdroppers are "silent," making CSI estimation more challenging, and to further explore the trade-offs between TS and PS in interference-limited environments.
Summary updated