Converging Time Synchronization Algorithm In Intelligent Transportation Systems

Intelligent Transportation System (ITS) aims to establish a real-time and effective traffic management system in which Advanced wireless communication, sensors and computer technologies are integrated in order to function as a whole. With the close harmony of people, vehicles and roads, ITS can sharply upgrade the transportation efficiency, improve traffic fluency as well as relieve traffic jam and environmental pollution. In ITS, Vehicular ad hoc networks (VANET) has played an essential role, which is a distributed and autonomous vehicle network.This dissertation describes a network time synchronization (NTS) algorithm for vehicular ad hoc networks (VANET), as well as traffic simulation in a target platform and evaluations of this algorithm in Matlab.The proposed time synchronization algorithm, referred to as Converging Time Synchronization (CTS), shows excellent scalability, accuracy and robustness against high mobility traffic scenarios. The CTS achieves its scalability by leading all vehicles synchronized to the existing largest synchronization group on roads. In order to further investigate the practical performance of CTS algorithm and compare it with other ones, we implement it in British Telecom.'s trafficCom platform. This platform is developed by professor Maziar Nevokee to simulate different traffic scenarios and implement our CTS algorithm. TrafficCom platform, characterized by discernable moving vehicles, is based on the Intelligent Driver Model (IDM). Served as a microscopic car-following model, IDM implements vehicle-vehicle interactions and tracks the motion of individual vehicles.Additional in our target platform, the implementations of both CTS and the Reference Broadcast Synchronization (RBS) algorithms from the literature are also presented, which contributes to improving and evaluating out proposed algorithm. After analyzing our results over Matlab, numerical evaluations demonstrate better performance of CTS algorithm against RBS algorithm under a high traffic flow scenario. Comparing with RBS algorithm, our CTS algorithm is more efficient on synchronization stability as well as synchronization range. Although a CTS process has two more broadcasts relative to the RBS process. But the larger synchronization range can compensate this time delay for CTS process. Furthermore, when a new vehicle comes, the synchronized group does not need to resynchronize, no matter how fast the clock of the coming vehicle is.Finally, the conclusions, further recommendations of our proposed algorithm as well as synchronization development in the target platform will be discussed. Future work will focus on the extension of our CTS algorithm. The algorithm parameters can be investigated to vary dynamically in response to changing traffic conditions. Furthermore, in CTS algorithm a series of broadcast-response steps have been defined and synchronization group scale has been taken as synchronizing priority. With these broadcast-response steps, we can abandon the scale and adopt other parameters as the synchronizing priority, e.g. latest updating time for accurate clock etc.