A Microservice-Based Industrial Control System Architecture Using Cloud and MEC.

Cloud computing has been adapted for various application areas. Several research projects are underway to migrate Industrial Control Systems (ICSs) to the public cloud. Some functions of ICSs require real-time processing that is difficult to migrate to the public cloud because network latency of the internet is unpredictable. Fog computing is a new computing paradigm that could address this latency issue. In particular, Multi-access Edge Computing (MEC) is a fog computing environment integrated with the 5G network, and therefore the real-time processing requirement of ICSs could be satisfied by using MEC. In this paper, we propose a microservice-based ICS architecture using the cloud and fog computing. In the architecture, each function of an ICS is implemented as a microservice and its execution locations are determined by an algorithm minimizing the total usage fee for cloud and fog computing while satisfying the real-time processing requirement. The proposed architecture and placement algorithm are evaluated by simulation under the scenario of a virtual power plant that manages distributed energy resources. The simulation result shows the proposed placement algorithm suppresses VM usage fee while satisfying the requirement of a real-time control function.