UC San Diego

The Internet of Things (IoT) has experienced remarkable growth in recent years, with the number of IoT devices reaching 11.3 billion by 2020, surpassing even the global population as well as the combined market of smartphones, tablets, and PCs. However, this growth has been slower than the previous predictions of trillions of deployed IoT devices within the past decade. One of the primary reasons for this slower growth is the challenges posed by existing battery-supported architecture, including high device and maintenance costs, as well as environmental concerns, all of which hinder scalability. To overcome these obstacles, there is a proposal for battery-free IoT devices that can harvest energy from ambient sources. However, The conventional active radios used in IoT devices consume tens to hundreds of milliwatts of power, making them unsuitable for energy harvesting, which typically provides less than 10 μW of power. In response, researchers have been exploring new radio architectures for ultra-low-power (ULP) communication and sensing.

However, ULP communication and sensing techniques face reliability challenges that hinder their practical deployment. Two specific challenges are identified: Firstly, widely adopted backscatter communication systems are susceptible to double attenuation of the two-part channel, making them vulnerable to blockages and environmental changes. Secondly, ULP sensing systems typically have low bandwidth, making them susceptible to issues in indoor multipath-rich environments.

To address the reliability problem, this dissertation proposes the following contributions: Firstly, it introduces a novel system architecture that enables micro-watt-level active transmission, thereby improving communication reliability. Additionally, the system adopts an asymmetric communication scheme to reuse commodity devices, enhancing practicality and efficiency. Secondly, the dissertation presents a long-range magnetic RFID system that utilizes magnetic signals instead of electromagnetic signals. This innovative approach helps reduce the impact of blockages and environmental factors, ensuring more reliable and consistent performance. Finally, the dissertation introduces a multi-antenna wideband UHF RFID localization system that leverages the frequency-agnostic property of backscatter to collect wide bandwidth RFID signals. This system achieves more accurate and dependable localization results, particularly in challenging multipath-rich indoor environments.