| Traffic volume measurement is critical in transportation engineering and vehicular networks. Existing research on traffic volume measurement mainly focuses on single-point traffic statistics. In this dissertation, we switch our view from single-point to multi-point , and study the important problem of privacy-preserving multi-point traffic volume measurement in intelligent cyber-physical road systems (CPRS), which complements the state of art. We take advantage of the capabilities provided by CPRS to exploit the potential for a fundamental shift in the way how traffic data in support of multi-point traffic volume measurement can be automatically collected. The objective is to allow transportation authorities to automatically collect and efficiently measure the aggregate multi-point traffic volume data from CPRS without learning information about individual vehicles.;In this dissertation, we start with the problem of privacy-preserving two-point traffic volume measurement in CPRS, and propose four novel measurement schemes to solve this problem, with varying degrees of efficiency, accuracy, and privacy. Our first two schemes protect vehicles' identities through keyed signatures based on a family of commutative one-way hash functions, and they can achieve exact measurement results. The third and fourth schemes achieve better privacy for vehicles through shared bit array masking, protecting vehicles' identities as well as their travelling trajectory. They are also much more efficient, and can gracefully control the tradeoff between vehicles' privacy and measurement accuracy. In particular, our third scheme utilizes fixed-length bit arrays, and it works great under the assumption of similar traffic among different locations. Our fourth scheme removes this assumption of traffic similarity through variable-length bit arrays, and it can fit in more realistic situations where different locations observe different traffic volume.;After that, we extend our idea of variable-length bit array masking to address the problem of privacy-preserving three-point traffic measurement, and eventually present a framework to deal with the general problem of privacy-preserving multi-point traffic measurement. We demonstrate the feasibility, scalability, and superior performance of our solutions through mathematical proofs, numerical analysis, as well as extensive simulations. The research results in this dissertation can be applied to a broad spectrum of applications in vehicular networks and transportation engineering. Furthermore, they have potential applications beyond vehicular networks, such as privacy-preserving traffic estimation in a subway system with tagged toll cards. It is also possible for them to be used for estimating the movement patterns of mobile users in a corporate wireless network. |
