Using Vehicular Networks for Urban Surveillance: an Adaptive Data Collection Scheme

In this paper we consider a vehicular sensor network in which vehicles are equipped with video cameras and continuously capture images from urban roads. Then, vehicles can use roadside wireless access points (APs) encountered during travel to deliver recorded image data to remote data collectors, where images streams from multiple sources are aggregated and processed. However, how to efficiently utilize the limited upload capacity of the wireless access network while reducing data redundancy due to spatial correlation of neighboring vehicles is a critical issue. To tackle this problem we propose a mechanism to dynamically adjust sampling rates of onboard cameras based on the vehicle status and the spatial distribution of roadside APs. The key idea is that vehicles traveling close to a roadside AP should use lower sampling rates than vehicles traveling in areas with a poor connectivity. We use network simulations with realistic vehicular mobility patterns to verify the performance gains of our proposed scheme compared to a baseline system that use fixed sampling rates. Simulation results show that our solution can ensure a more balanced and uniform coverage of the road network while reducing the amount of transferred data.