Ant-based vehicle congestion avoidance framework using vehicular networks / Mohammad Reza Jabbarpour Sattari

Sattari, Mohammad Reza Jabbarpour (2015) Ant-based vehicle congestion avoidance framework using vehicular networks / Mohammad Reza Jabbarpour Sattari. PhD thesis, University of Malaya.

Preview

PDF (Full Text) Download (4Mb) | Preview

Abstract

Over the last decade, vehicle population has dramatically increased all over the world. This large number of vehicles coupled with the limited capacity of the roads and highways lead to heavy traffic congestion. Besides, it gives rise to air pollution, driver frustration, and costs billions of dollars annually in fuel consumption. Although finding a proper solution for vehicle congestion is a necessity, it is still remaining a challenging task due to the dynamic and unpredictable nature of vehicular environments. Building new high-capacity streets can be a solution but it is very costly, time consuming and in most cases, infeasible due to space limitations. However, optimal usage of the existing roads and streets capacity can lessen the congestion problem in large cities at a lower cost. Intelligent Transportation System (ITS) is a newly emerged system that aims to provide innovative services for different modes of transportation and traffic management. Vehicle Traffic Routing System (VTRS) is one of the ITS applications that can be used for efficient utilization of existing roads’ capacity. Previous researches concentrated on using static algorithms to find the shortest path in VTRSs. However, providing a shortest path without considering other factors such as congestion, accidents, obstacles, travel time and speed is not a proper solution for vehicle traffic congestion problem. The efficiency of VTRSs on mitigating the vehicle congestion is challenged by the high dynamicity and quick changes of vehicular environments due to both predictable (recurring) and unpredictable (non-recurring) events. Most of the existing approaches deal with the congestion problem in a reactive manner and recover vehicle congestion implicitly, which is not a sufficient solution due to non-recurring congestion conditions. Moreover, a same path is suggested to drivers by the existing approaches which switches the congestion from one route to another, specifically, in the case of having a significant number of drivers utilizing these systems simultaneously. This research presents a bio-inspired framework, called “Ant-based Vehicle Congestion Avoidance Framework (AVCAF)”, which is a promising way to alleviate vehicle traffic congestion problem while considering the aforementioned drawbacks. AVCAF predicts vehicles’ average travel speed and combines it with travel time, density, distance, map segmentation and layering to reduce congestion as much as possible by finding the least congested shortest paths in order to avoid congestion instead of recovering from it. AVCAF uses alternative paths from the early stages of the routing process. AVCAF collects real-time traffic data through vehicular networks to consider non-recurring congestion conditions in its routing mechanism via ant-based algorithm. The proposed framework is evaluated and validated through simulation environment. Experimental results conducted on three different scenarios (i.e. various vehicle densities, various system usage rates and accident condition) considering average travel time, speed, distance, number of re-routings and number of congested roads as evaluation metrics. The results show that AVCAF outperforms the existing approaches in terms of average travel time, travel speed, number of re-routings and number of congested roads.

Actions (For repository staff only : Login required)