Research of Clustering Methods of ARM Devices Based for Edge Computing Use Cases

The work is devoted to the study of clustering methods for ARM devices to organize data processing and computing in a local network for edge computing use cases. The paper presented an analysis of the subject area, containing the study of logical levels of edge computing, the study of container virtualization technologies and clustering methods that could be applied to devices used in edge computing. The study of clustering methods revealed that Kubernetes-based solutions are the most user and developer friendly and provide the greatest number of functional and non-functional features. The best solution according to the given criteria was identified, its advantages and disadvantages were analyzed. The performance of the solution was tested and benchmarked against its public cloud counterparts. The test results showed that the performance of the solution is slightly inferior to their counterparts deployed in cloud providers.