Dynamic Vehicle Re-routing to Avoid Congested Roads

Abstract

As the traffic congestion increases, vehicles can change their current path aiming at minimizing the travel time.Dynamic path planning using real-time traffic information have attracted the interest of both academia and industry recently especially using vehicular adhoc-networks(VANETs). An important aspect of the path planning using VANETs is the traffic information collection. To collect traffic information a vehicle initiates number of queries, which are routed by VANETs along different paths toward its destination. However the driver may not need the information of routes with higher travel times. Existing works have addressed the traffic information collection problem in several ways. However they have certain weaknesses on efficiency and effectiveness in the data collection approach. Few works have focused on optimizing and filtering un-useful traffic information collection queries. This dissertation introduces several filtering algorithms to optimize the traffic data collection process. Using these filtering techniques intermediate vehicles will be able discard some of the data collection queries before they reach the source vehicle. By discarding queries very early communication cost of the vehicular network can be minimized. A Simulator based performance evaluation was carried out to evaluate the performance of the proposed algorithm. Calculations have been done based on a real data set. Results suggests that communication cost can be significantly reduced when the proposed filtering techniques were applied.