Value-of-Information Driven Communication for Connected and Automated Transportation (VoiiComm)

Research Project, 2020 – 2022

By fusing the sensing information provided by the automated vehicles (AVs), an accurate and real-time representation of the highly dynamic traffic environment can be formed and disseminated to all traffic participants, to enable a variety of safety and efficiency improving applications and push the transition towards connected and automated transportation. A powerful wireless network plays a key role in this vision. As the number of AVs increases, however, the huge data volume will easily exceed the capacity of the wireless network. Owing to the application-oriented nature of vehicular communication, a paradigm shift towards the value-of-information (VoI) driven design is crucial to avoid wasting network resources and degrading traffic safety and efficiency. Nevertheless, no simple mathematical model and existing theory can be applied to assess VoI exactly.

The main goal of the VoiiComm project is to establish a theoretical framework and practical methods for evaluating and modeling VoI statistically, by exploiting the intrinsic relationships between traffic flow, density, and speed, to enable the systematic design of efficient communication protocols. The specific research objectives are (1) Design metrics that capture the space-time statistical properties of VoI; (2) Analytical and empirical modeling of spatio-temporal VoI metrics; (3) Strategy development for efficient sampling and transmission scheduling. By accomplishing these objectives, data congestion and resource waste can be avoided, and the potential benefits of AVs can be maximized as their number gets large. VoiiComm will initiate a fresh new interdisciplinary research area and advance the transformation to sustainable and automated transportation and society. In parallel, it will allow the ER to achieve several knowledge transfer and training objectives, increase prominence in her research field, and enhance her career prospects in academia.