Abstract
Energy beamforming has emerged as a promising technique for enhancing the energy transfer efficiency of wireless power transfer (WPT). However, the performance of conventional energy beamforming may seriously degrade due to the non-linear radio frequency (RF) to direct current (DC) conversion at energy receivers (ERs). To tackle this issue, this letter proposes a new time-division energy beamforming, in which different energy beamforming matrices (of high ranks in general) are time shared to exploit the "convex-concave" shape of the RF-DC power relation at ERs. By considering a particular time duration for WPT, we maximize the minimum harvested DC energy among all ERs, by jointly optimizing the energy beamforming matrices and the corresponding time allocation. In order to solve the non-convex min-DC-energy maximization problem, we propose an efficient solution by using the techniques of alternating optimization and successive convex approximation (SCA). Numerical results show that the proposed time-division energy beamforming design indeed outperforms the conventional multi-beam and time-division-multiple-access (TDMA)-based energy transmissions.
Introduction
In multiuser Wireless Power Transfer (WPT), an energy transmitter (ET) equipped with multiple antennas aims to deliver power to several energy receivers (ERs). Traditional research often relies on a linear energy harvesting (EH) model, assuming harvested DC power is directly proportional to received RF power. However, practical rectifiers exhibit significant non-linearity. The RF-to-DC conversion follows a "convex-concave" curve: a convex region at low power where efficiency is poor, and a concave region at high power where saturation occurs.
As shown in Figure 1, the ET uses $M$ antennas to transmit energy to $K$ ERs. Each ER consists of a rectifier that converts the RF signal to a DC signal to charge a battery. When users are in the low-power convex regime, standard "multi-beam" strategies that distribute power continuously often fail to achieve high conversion efficiency because they "spread the butter too thin."
Proposed Time-Division Energy Beamforming
To overcome the limitations of static beamforming, this paper proposes a Time-Division Energy Beamforming framework. The core idea is to divide the transmission time into multiple slots, allowing the ET to switch between different beamforming matrices. This "pulsing" effect can push receivers into the more efficient concave region of their harvesting curves.
Optimization Framework
The approach utilizes a sigmoid-based EH model to accurately capture hardware non-linearities. The optimization objective is to maximize the minimum harvested DC energy across all ERs to ensure fairness. The resulting non-convex problem is solved using:
- Alternating Optimization (AO): Iteratively optimizing beamforming shapes and time slot durations.
- Successive Convex Approximation (SCA): Using Taylor expansions to linearize the difficult sigmoid functions into manageable convex sub-problems.
Results and Discussion
Numerical simulations using Rician fading channels demonstrate the superiority of the proposed method. A key finding is that in low-power regimes, the proposed design significantly outperforms conventional methods.
As illustrated in Figure 2, at a transmit power of 38 dBm, the proposed time-division energy beamforming achieves approximately a 7% gain in harvested DC power over the multi-beam approach. Furthermore, the proposed method consistently outperforms pure TDMA-based transmission. The strategy is adaptive: it converges to multi-beam transmission in high-power regimes (concave region) and behaves like TDMA in extremely low-power regimes, providing a unified optimal framework.
Conclusion
This work highlights the importance of accounting for non-linear hardware characteristics in WPT design. By employing time-division energy beamforming, significant efficiency gains can be realized, especially for users receiving low RF power. Future research will address the challenges of obtaining perfect Channel State Information (CSI) and managing the hardware complexity of rapid beam switching in dynamic environments.
Summary updated