Impact of Resource Distribution on Performance of Fog Computing Infrastructure

Abstract

The data being produced by the devices connected to the Internet is ever growing. The large volume of data is also required to be processed and stored. Cloud computing provides data processing and storage services by using IT resources at gegraphically distributed locations. However, the cloud computing resources are usually far away from the sites where data is produced and processing is needed, which results in communication latencies that may not be acceptable for some appliations. This issue is addressed by Fog computing by brining data processing closer to the edge of the Internet. As fog infrastructures are significantly different from cloud in terms of scale, processing capacity, bandwidth etc.; more focused analysis is needed to explore avenues for performance improvement. In this paper, we analyze effects of resource distribution in small-scale fog computing infrastructure on its performance and resource utilization. We examine different topological arrangements of fog devices with varying resources. Our results show that the topological arrangement of fog nodes has a significant impact on the overall performance of fog computing infrastructure.