Scalable Indoor Localization Using Ultra-Wideband Radios

The recent advances in the Internet of Things (IoT) technologies have started a new era in sensor networks and smart assistant systems. Various types of sensing platforms are being deployed to understand the in-depth behavior of people while maintaining human comfort. Technology that tracks people inside buildings could become a key enabler for many applications in this space. Indoor localization is a process to find the exact location of devices, objects or people inside buildings in which GPS service is mostly unreliable. Existing indoor localization and tracking solutions can be divided into two main categories: passive and active solutions. Passive asset tracking systems are scalable, but their accuracy is limited to a few meters (Room Level). On the other hand, in active tracking scenarios, the target has to carry a tacking device, which makes the location estimation more accurate and robust. In this dissertation, we improve the scalability and robustness of indoor tracking solutions. Ultrawideband (UWB)-based indoor localization techniques are one of the well-known and popular active indoor tracking systems. Large bandwidth of UWB signals makes them resilient to multipath fading problem and brings the ability to estimate the location of a target with a few centimeters error. Despite the recent advancement of the accuracy of UWB based indoor tracking systems, the scalability of these systems did not receive enough attention from the research community until the last few years. In this dissertation, we focus on four primary challenges in scalability of UWB systems: adaptively finding optimum UWB physical layer setting to achieve best ranging performance while maintaining application requirements, reducing deployment constraints by proposing single anchor UWB indoor localization, studying and mitigating the impact of multi-user interference on UWB ranging, and combining ranging traffic with non-ranging traffic to increase the applicability of UWB networks for non-ranging applications.