MASSACHUSETTS INSTITUTE OF TECHNOLOGY

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5). The coming years will see the emergence of location-aware applications with sub-meter accuracy and minimal infrastructure requirements, operational in harsh environments. This project seeks to gain a fundamental understanding of the theoretical and practical benefits of cooperative localization and tracking, wherein mobile devices are able to infer their instantaneous positional information. The research is motivated by the need for positional information in numerous commercial, public service, and military applications. Current localization technologies (e.g., GPS and Wi-Fi) are not suitable for these applications because they do not provide the required accuracy and robustness. In cooperative localization and tracking, network devices work together to determine positional information. In contrast to non-cooperative technologies, cooperative tracking is provably more accurate and more robust. Despite recent research on cooperative localization and multi-lateration, there are still several key challenges that impede the application to real-world situations. The project will address the following critical questions: (1) How good are current algorithms, and can we do significantly better? Many of the fundamental aspects of cooperative tracking are not well understood, and it is thus unclear if substantial performance gains are achievable. (2) Do theories of cooperative localization and tracking translate to the real-world environment? To the best of the PI's knowledge, wideband cooperative localization networks have not been demonstrated in realistic environments. (3) Can we improve ranging techniques to work in the presence of physical obstructions? Non-line-of-sight (NLOS) propagation conditions due to obstructions cause significant errors in range estimates. To address these challenges, the project has the following objectives: - Determine ultimate theoretical limits and bounds on cooperative tracking networks. These limits will offer insight into the performance versus complexity trade-off. Once the limits are found, the PI will then develop cooperative, distributed algorithms that perform close to these limits. - Design and implement an ultrawide bandwidth (UWB) radio network that demonstrates cooperative localization and tracking. The design and development of the network will involve many undergraduate students, thus offering them unique exposure to aspects of wireless networks. - Develop algorithms to detect NLOS conditions to further increase localization performance. These algorithms will be based on experimental data collected using a UWB testbed. The proposed research will take large strides toward the ubiquitous deployment of robust and distributed tracking systems. The intellectual merit of this research includes the potential to advance the understanding of key issues of location-aware networks. This interdisciplinary project consists of both theoretical and experimental components. This offers the potential to lead to novel methods and techniques to solve self-tracking problems in a wider range of environments and under more stringent requirements compared to what is currently available. The potential broader impacts of this project include commercialization and dissemination of this technology. The project has already engaged local and state government agencies. The research results are integrated into graduate courses at MIT, as well as in tutorials and short courses offered elsewhere. Furthermore, results and measurement data are to be made publicly available via the Internet. To advocate diversity and outreach, the project is used to host students from both MIT's Undergraduate Research Opportunities Program (UROP) and MIT's Summer Research Program (MSRP).

Clarification of Codes