Simulation time experiments in distributed traffic simulation environments

Abstract

Transportation system management and traveler information systems evolve with the development of data communications and traffic simulations. Variety of roadside and mobile sensing platforms will be deployed to allow communication between vehicles with Dedicated Short Range Communications (DSRC). Traffic data received from moving vehicles will be transmitted to each individual vehicle and traffic management center to provide real time traffic information. Microscopic traffic simulation models will be used for traffic analysis and prediction, since they have the highest detailed level of prediction such as vehicle /driver characteristics and have the capability to capture dynamically changing traffic conditions through the simulation. In this study, three communication methods are used including Ethernet, off-the-shelf wireless network, and one commercial network provider for data communication. Simulation time is measured using three different communication methods and one non-communication case. Based on the Kruskal-Wallis test and multiple comparisons using Bonferroni Adjustment, Fisher's Protected LSD and Tuckey's HSD, it can be concluded that simulations with off-the-shelf wireless network or 'Commercial network provider' are different with 'no communication' or 'Ethernet communication'. Also, the original signal plan and new signal plan do not make a significant difference on travel delay. © 2014 IEEE.