Adaptive resource allocation based on the billing granularity in edge-cloud architecture

With the rapid development and popularization of the Internet of Things technology, the number of network edge devices has increased rapidly, and the Internet of Things perception layer has generated massive data. Cloud-Edge collaboration will be used more and more to solve this problem. When the business reaches its peak, cloud computing capabilities cannot meet business needs. The cloud service provider can apply for resources to meet the requirements of the computing resources. After the peak period of the business, if there are idle computing resources, the resources can be released to the cloud service provider, which can reduce the service cost and save the computing resources. The traditional flexible resource management is mainly used in a single cloud environment, and the prediction of the resource demand is insufficient. The factor of the cloud resource billing granularity is neglected, resulting in high cloud lease cost. Therefore, an adaptive resource allocation algorithm and a data migration algorithm are proposed. The prediction algorithm provides the basis for the adaptive resource allocation of the edge cloud cluster. The adaptive resource allocation algorithm determines the resource allocation scheme of the edge cloud cluster with the lowest service cost. The data migration algorithm guarantees the reliability of data and achieves cluster load balancing. A large number of experimental results show that our newly proposed algorithm can greatly improve system performance in terms of better cost control, higher data integrity and load balancing.