Density Adaptive Urban Data Collection in Vehicular Sensor Networks

Research on vehicular sensor networks (VSNs) has drawn increasing attentions recently. By collecting data from onboard sensors and gathering them with VSNs, vehicles can serve as data sources for many applications such as intelligent transportation system, smart cities and etc. Existing data collection solutions in VSNs focused on temporal or spacial influence but did not address the influence of vehicle density. In this paper, we present a novel data collection solution for collecting data within a specific geographical region in urban scenario. The objective of proposed algorithm is to satisfy the specific requirement of applications by collecting certain number or ratio of data. It is able to adjust the required collection life cycle adaptively according to vehicle density. The mobility nature of vehicles determines that the topology of VSN is dynamic. To adapt with mobility, the algorithm is designed to be decentralized and stateless. Therefore, vehicles implementing this algorithm do not require neighboring information or link state. The network transmission latency is analytically characterized and modeled. Evaluations show that the results of the algorithm match the mathematical model well and outperforms existing protocols in terms of efficiency and effectiveness.