RF Wireless Power and Data Transfer: Experiment-driven Analysis and Waveform Design

The brisk deployment of the fifth generation (5G) mobile technology across the globe has accelerated the adoption of Internet of Things (IoT) networks. While 5G provides the necessary bandwidth and latency to connect the trillions of IoT sensors to the internet, the challenge of powering such a multitude of sensors with a replenishable energy source remains. Far-field radio frequency (RF) wireless power transfer (WPT) is a promising technology to address this issue. Conventionally, the RF WPT concepts have been deemed inadequate to deliver wireless power due to the undeniably huge over-the-air propagation losses. Nonetheless, the radical decline in the energy requirement of simple sensing and computing devices over the last few decades has rekindled the interest in RF WPT as a feasible solution for wireless power delivery to IoT sensors.

The primary goal in any RF WPT system is to maximize the harvested direct current (DC) power from the minuscule incident RF power. As a result, optimizing the receiver power efficiency is pivotal for an RF WPT system. On similar lines, it is essential to minimize the power losses at the transmitter in order to achieve a sustainable and economically viable RF WPT system. In this regard, this thesis explores the system-level study of an RF WPT system using a digital radio transmitter for applications where alternative analog transmit circuits are impractical. A prototype test-bed comprising low-cost software-defined radio (SDR) transmitter and an off-the-shelf RF energy-harvesting (EH) receiver is developed to experimentally analyze the impact of clipping and nonlinear amplification at the digital radio transmitter on digital baseband waveform. The use of an SDR allows leveraging the test-bed for the research on RF simultaneous wireless information and power transfer (SWIPT); the true potential of this technology can be realized by utilizing the RF spectrum to transport data and power together. The experimental results indicate that a digital radio severely distorts high peak-to-average power ratio (PAPR) signals, thereby reducing their average output power and rendering them futile for RF WPT.

On similar lines, another test-bed is developed to assess the impact of different waveforms, input impedance mismatch, incident RF power, and load on the receiver power efficiency of an RF WPT system. The experimental results provide the foundation and notion to develop a novel mathematical model for an RF EH receiver. The parametric model relates the harvested DC power to the power distribution of the envelope signal of the incident waveform, which is characterized by the amplitude, phase and frequency of the baseband waveform. The novel receiver model is independent of the receiver circuit’s matching network, rectifier configuration, number of diodes, load as well as input frequency. The efficacy of the model in accurately predicting the output DC power for any given power-level distribution is verified experimentally.

Since the novel receiver model associates the output DC power to the parameters of the incident waveform, it is further leveraged to design optimal transmit wave-forms for RF WPT and SWIPT. The optimization problem reveals that a constant envelope signal with varying duty cycle is optimal for maximizing the harvested DC power. Consequently, a pulsed RF waveform is optimal for RF WPT, whereas a continuous phase modulated pulsed RF signal is suitable for RF SWIPT. The superior WPT performance of pulsed RF waveforms over multisine signals is demonstrated experimentally. Similarly, the pulsed phase-shift keying (PSK) signals exhibit superior receiver power efficiency than other communication signals. Nonetheless, varying the duty-cycle of pulsed PSK waveform leads to an efficiency—throughput trade-off in RF SWIPT.

Finally, the SDR test-bed is used to evaluate the overall end-to-end power efficiency of different digital baseband waveforms through wireless measurements. The results indicate a 4-PSK modulated signal to be suitable for RF WPT considering the overall power efficiency of the system. The corresponding transmitter, channel and receiver power efficiencies are evaluated as well. The results demonstrate the transmitter power efficiency to be lower than the receiver power efficiency.