TWC: Small: Evidence of Presence for Intelligent Vehicles using Environment-Based Security

点击翻译按钮获取中文摘要

Emerging intelligent automobiles will be able to harness advance on-car sensors to support new applications such as pollution detection, road condition monitoring, and traffic control. All these applications require the ability to verify both the location and the time of a reading. This project involves the design of verification methods that make use of environment factors, such as the presence of light and shadows and the measured wireless signal strength, instead of conventional public key infrastructure-based methods, in order to verify when and where data was collected. This new environment-based paradigm is resilient against insider attacks, easier to deploy, and protects the individual car owner's privacy. This research will help in realizing applications arising from the increasing presence of smarter vehicles on our roads which can benefit the public's wellbeing. The project also incorporates outreach components for high school and college students through the organization of science competitions, national undergraduate workshops, and summer research camp activities. This project is the first effort to use the wireless communication and video recording abilities of modern automobiles to capture natural environment characteristics to securely verify spatial-temporal claims in a vehicular network setting. This is an interdisciplinary research effort combining wireless networking and computer vision to address vehicular network security. The main project goals are: (1) explore new vision analytic algorithms in order to identify the location and time images are captured by an automobile camera; (2) research new algorithms for identifying optimal roadside unit locations for location disambiguation to support wireless spatial-temporal verification; (3) develop new techniques for utilizing encounters with public vehicles for verifying spatial-temporal claims; (4) developing a fusion framework to combine wireless measurements, vehicular encounters, and visual images for spatial-temporal verification; and (5) perform realistic experiments on real roads and vehicular test bed to collect image and wireless datasets and evaluate the research. These datasets are of interest to both wireless networking and computer vision research communities, and will be made available to the public.