Return to search

Vehicular Cloud: Stochastic Analysis of Computing Resources in a Road Segment

Intelligent transportation systems aim to provide innovative applications and services relating to traffic management and enable ease of access to information for various system users. The intent to utilize the excessive on-board resources in the transportation system, along with the latest computing resource management technology in conventional clouds, has cultivated the concept of the Vehicular Cloud. Evolved from Vehicular Networks, the vehicular cloud can be formed by vehicles autonomously, and provides a large number of applications and services that can benefit the entire transportation system, as well as drivers, passengers, and pedestrians. However, due to high traffic mobility, the vehicular cloud is built on dynamic physical resources; as a result, it experiences several inherent challenges, which increase the complexity of its implementations.
Having a detailed picture of the number of vehicles, as well as their time of availability in a given region through a model, works as a critical stepping stone for enabling vehicular clouds, as well as any other system involving vehicles moving over the traffic network. The number of vehicles represents the amount of computation capabilities available in this region and the navigation time indicates the period of validity for a specific compute node. Therefore, in this thesis, we carry out a comprehensive stochastic analysis of several traffic characteristics related to the implementation of vehicular cloud inside a road segment by adopting proper traffic models. According to the analytical results, we demonstrate the feasibility of running a certain class of applications or services on the vehicular cloud, even for highly dynamic scenarios.
Specifically, two kinds of traffic scenarios are modeled: free-flow traffic and queueing-up traffic. We use a macroscopic traffic model to investigate the free-flow traffic and analyze the features such as traffic density, the number of vehicles and their residence time. Also, we utilize the queueing theory to model the queueing-up traffic; the queue length and the waiting time in the queue are analyzed. The results show the boundaries on enabling vehicular cloud, allowing to determine a range of parameters for simulating vehicular clouds.

Identiferoai:union.ndltd.org:uottawa.ca/oai:ruor.uottawa.ca:10393/34793
Date January 2016
CreatorsZhang, Tao
ContributorsBoukerche, Azzedine
PublisherUniversité d'Ottawa / University of Ottawa
Source SetsUniversité d’Ottawa
LanguageEnglish
Detected LanguageEnglish
TypeThesis

Page generated in 0.0022 seconds