Search

There are billions of Internet of Things (IoT) devices distributed across the globe, and this growing number of interconnected IoT devices demand seamless networking and low-power communication. While many devices are powered with batteries, their limitations such as maintenance and environmental impact call for battery-free alternatives. Small battery-free devices are attractive for sensing as they can use backscatter communication and operate on harvested energy from their surroundings. This dissertation presents a collection of novel techniques for backscatter communication, a method that reduces energy consumption by several orders of magnitude compared to standard low-power radio communication. Backscatter communication provides a direction for implementing widespread networks of battery-free devices that can be used for ubiquitous sensing. However, real-world deployment of backscatter tags encounters challenges due to their constrained power budgets. Adding mechanisms for identification, scheduling, querying and relaying for backscatter should be done carefully offloading power consuming components and delegating tasks whenever possible to an external powerful device. This dissertation advances the state of the art in two different kinds of backscatter networks: digital backscatter networks and analog backscatter networks. Like conventional RF devices, protocol-based digital backscatter tags encode and communicate binary data in packets, allowing these tags to interoperate with conventional IoT devices using protocols such as IEEE 802.15.4. Applications such as dense networks require tag-to-tag multi-hop communication which introduces challenges as the tags rely on an external signal. For digital backscatter, I present protocol-based multi-hop communication and develop a tool to test large tag-to-tag networks. By contrast, analog backscatter directly communicates the sensor readings by modulating the external signal. As the analog tags lack a packet structure an