There is a rapidly growing interest in the use of cloud computing forautomotive vehicles to facilitate computation and data intensive tasks.Efficient utilization of on-demand cloud resources holds a significantpotential to improve future vehicle safety, comfort, and fuel economy. In themeanwhile, issues like cyber security and resource allocation pose greatchallenges. In this paper, we treat the resource allocation problem forcloud-based automotive systems. Both private and public cloud paradigms areconsidered where a private cloud provides an internal, company-owned internetservice dedicated to its own vehicles while a public cloud serves allsubscribed vehicles. This paper establishes comprehensive models of cloudresource provisioning for both private and public cloud- based automotivesystems. Complications such as stochastic communication delays and taskdeadlines are explicitly considered. In particular, a centralized resourceprovisioning model is developed for private cloud and chance constrainedoptimization is exploited to utilize the cloud resources for best Quality ofServices. On the other hand, a decentralized auction-based model is developedfor public cloud and reinforcement learning is employed to obtain an optimalbidding policy for a "selfish" agent. Numerical examples are presented toillustrate the effectiveness of the developed techniques.
展开▼
