Design of Scheduling Algorithms for Vehicular Cloud Computing

Abstract

In the last decade, vehicular cloud computing has received huge attention in business and scientiand#64257;c commu- newlinenities. The domain of vehicular cloud computing integrates two emerging and#64257;elds, namely cloud computing newlineand vehicular ad hoc networks. It acts as a data center by using the underutilized resources of the networked newlinevehicles. Moreover, many studies suggest that these vehicles are potential candidates for hosting virtual ma- newlinechines (VMs). As a result, a VM can be set up using the vehicular resources, and it can be transferred from newlineone vehicle to another vehicle to continue its execution, under some circumstances. It enables the vehicular newlineenvironment to provide services to the user requests (URs), that are submitted to the cloud, by the users. newlineHowever, the mapping of such requests to the VMs (or hosted vehicles) and service migration is very much newlinechallenging, and not well-studied in the literature. newlineThis thesis s elementary focuses are addressing the various challenges of the vehicular cloud environ- newlinement and developing scheduling and migration algorithms. Here, we consider various inand#64258;uencing factors newlinelike speed, parking time, mobility pattern, vehicles energy source, resource utilization, energy consump- newlinetion, and cost in the algorithms development process to achieve better performance. In the beginning, we newlinepropose a dynamic service migration (DSM) algorithm for vehicular clouds. The algorithm consists of three newlinephases, estimation, assignment and migration. The performance analysis is carried out through simulation newlineusing two scenarios of six datasets, and compared with three well-known algorithms, namely vehicular VM newlinemigration-uniform (VVMM-U), round robin and mobility and destination workload aware migration (MD- newlineWLAM) using four performance measures. The comparison results followed by statistical validation using newlineT test show the superiority of the proposed algorithm over the existing algorithms. newlineNext, we propose a smart cloud service management (SCSM) algorithm for vehicular clouds (VCs) and newlineaddress numerous cha