Dynamic Resource Controller Technology to Accelerate Processing and Utilization of IoT Data

Recently, a concept called edge computing has been attracting attention for full-scale application of the Internet of Things (IoT) to business, and companies have started working on its use. Edge computing is intended to improve a computing system's response to terminals by distributing data processing between nodes provided near the site, rather than having all data processed in a cloud. It is hoped this will restrain the huge increase in the volume of communication traffic produced by the IoT. Fujitsu Laboratories has extended this concept and developed dynamic resource controller (DRC) technology, which utilizes computers (edge nodes) installed near the terminal on the site as if they are integrated into the cloud. This technology dynamically optimizes the locations of data processing and storage between the edge nodes and cloud according to changes in the site environment. Unlike the conventional edge computing, this not only reduces the usage of network resources but also allows for a positive increase in the usage of any surplus network bandwidth to gather raw data to the cloud that may lead to new discoveries. It can also be used to deal with addition of devices and edge node load variation by changing the location of some of the processing to edge nodes with surplus resources, which helps maintain the stability of the system. This paper describes the mechanism of the DRC technology, which allows a computing system to make a quick decision on the appropriate location of processing execution to deal with addition of devices and variation of the edge node load, and presents the results of its evaluation.