Real Time Traffic Congestion Prediction Approach on a Road System Using Dynamic Clustering of Inter-vehicle Communication

Abstract

In the past two decades, the proliferation of new technology has made a huge impact in the life style of the people. Here transport is one of the fascinating technology branches developed during last 200 years and time is the very precious demand in today’s fast moving life. As more or more vehicles are being introduced into the populated countries like India, traffic congestion has become a serious problem in the context of transportation system. Due to this serious problem the overall transportation service become disturb and inconvenience in the crucial perspective of time. Accurate real time prediction of traffic congestion can therefore facilitate adaptive decision making process of both operations and passengers. Providing high quality information with time for generating better mobility and increased overall satisfaction with public transmit. Actually in our daily life situation, time assessment is needed for handling different travelling hazards. People can do any type of real time planning and scheduling by their thinking in their own desire. Even sometimes, feeling comfort and always getting updated time to time news is recommended as an enjoyable mind free journey. That’s why a successful implementation of travelling system will provide the passengers with an easy, reliable and time saving way of travelling and that will also greatly encourage the use of public transmit.

Infrastructure based existing solutions of traffic congestion prediction has several drawbacks with respect of time, cost, bandwidth and coverage area. So, here we adopt vehicular communication that can be used to predict and transmit traffic information. Clustering is one of the efficient transmission strategies in vehicular Ad-hoc network (VANET). It is an important approach in the process of geographic region formation for vehicular Ad-hoc network (VANET). In this paper, for real time traffic congestion prediction purpose we propose a dynamic inter-vehicle cluster formation strategy that is necessary for knowing how many vehicles are communicated with each other in any time instant and identify the regions in which vehicles are densely concentrated. Inter-vehicle communication systems rely on direct communication between running vehicles to satisfy the communication needs of large class of applications. Each vehicle works as a source, destination and/or a router to disseminate traffic related information. Each vehicle broadcast beacon message containing information about their location, speed, moving direction periodically to the other vehicles through light weight communication protocol. Inter-vehicle communication envisioned to provide broad range of applications in the area of traffic congestion information. So, for real time interaction between passengers is necessary for the implementation of real time analysis. The inter-vehicle cluster is a group of vehicles with similar mobility pattern as per inter-distance, speed and moving direction and also elects a leader, called cluster head, for each group. Cluster head have the responsibility to collect all the vehicles information within its own cluster area through inter-vehicle communication system and forward the merged result to the next cluster head. All the calculations inside the clusters are unbiased; all vehicles get the same priority for exchange their information to each other. In this way global traffic information about the whole road can be in priority propagated to each vehicle and all its passengers. So, here our intention is to develop a real time traffic congestion prediction approach on a road system in a very smart way for which time related any obstacles will be easily managed in any capacity to meet the desired demand.