Towards Job Completion Time in Vehicular Cloud by Overcoming Resource Volatility

Abstract

Future vehicles are expected to generate large amounts of data which may need to be off-loaded to a proximate server for processing. This led to the introduction of vehicular clouds (VC), which proposes that computing is done at nearby vehicles. However, as the vehicles may leave and join the VC randomly, the computing services of VC are time-varying, which may cause service interruptions. This work analytically evaluates the performance of the VCs under a service strategy that overcomes the interruptions caused by resource volatility. We use order statistics to derive the probability distribution of the number of vehicle arrivals to assign all the tasks of a job, the upper and lower bounds of mean job completion time, and the probability density function of the completion time of the longest task. Finally, we present the numerical results for the analysis and the simulation results to show the correctness of the analysis.