Research On The Key Techniques Of Vehicular Sensor Networks And Applications

As a new paradigm of Wireless Sensor Networks (WSN), Vehicular Sensor Network (VSN) attracted more and more attention from industry and academy. VSN is a network of mobile sensors equipped on vehicles and can be used for different applications, such as safe driving, real-time traffic monitoring and urban mobile surveillance. Compared with traditional WSN, VSN has its own characteristics, such as no energy constraints, high node density, dynamic network topology and passive coverage. As the new applications and technique issues mentioned above, this study focuses on the key techniques of vehicular sensor networks and applications. Specifically, we consider the following three issues: 1) information possessing in real-time traffic monitoring application; 2) DTN routing protocol design in VSN; 3) information cooperative storage in urban mobile surveillance application. Major contributions and the originalities of this dissertation are as follows:(1) The feasibility of vehicular sensor networks for traffic monitoring has been demonstrated and accurate map-matching and traffic status estimation algorithms have been proposed. This work can provide reference and experience for real deployment of such application.This study utilizes vehicular sensor networks of taxi companies for urban traffic monitoring. Unfortunately, previous approaches are not applicable because of low sampling rate of sensors and sparse sensory data, lossy information makes accurate traffic status estimation a real challenge. Based on such characteristics of data basis, effective map-matching and traffic status estimation algorithms have been proposed and then large-scale filed testing has been carried out. The testing results show that the algorithms can calculated real-time traffic status accurately by utilizing lossy sensory data, which demonstrates the feasibility of vehicular sensor networks for traffic monitoring.(2) Two DTN routing protocols for VSN have been proposed, which are designed by considering the features of VSN and have better routing performance. We need to design effective DTN routing protocol when deploying delay-tolerant applications in VSN. First, this study proposed a DTN routing protocol for VSN: POR protocol, which emphasizes neighbor selection based on the utilities of data packet and neighbor node. POR formulates DTN routing protocol as a utility optimization problem. The simulation results show POR outperforms traditional DTN routing protocols in VSN. Next, this study proposed a special DTN routing protocol for Bus Network (BusNet) by utilizing more information, such as bus schedule and bus line, to improve routing performance.(3) A cooperative storage system in VSN for mobile surveillance has been proposed, which is designed to maximize the average lifetime of sensory data for future event investigations.Currently, with the inclusion of forward facing cameras mounted on vehicles, there is an increasing interest in proactive urban sensing services where vehicles continuously capture images from urban streets. It can assist the reconstruction of crimes and, more generally, the forensic investigations of events monitored by VSN. Different from conventional sensor networks where data is dispatched to sinks, the node maintains the sensory data in their local storage for communication cost saving. This architecture can be also regarded as an in-network distributed storage system. This study presents VStore, a cooperative storage system in vehicular sensor networks for mobile surveillance. VStore is designed to maximize the average lifetime of image data in the network against data redundancy and storage imbalance between vehicles.