A Multi-layered Distributed Cloud Network for Cyber-Physical Energy System

Abstract

The diversity of power sources for electricity generation excites the current smart grid to evolve towards Cyber-Physical Energy System (CPES), which integrates with other systems, such as natural gas and heat. By advanced communication technologies, CPES can achieve effective energy fusion, real-time computation and precise control. With the increasing number of energy applications that bring large quantities of data, it calls for sufficient computing and storage resources with fast transmission and powerful processing to satisfy users' service level agreements (SLAs). This paper presents a multi-layered distributed cloud network (MDC) for CPES, with the design of the physical integration of energy systems, a distributed cloud architecture and a hierarchal information structure based on the fifth generation (5G). Furthermore, we present a two-level resource allocation model for cloud network, which aims to decide the location of deploying cloud facilities and dynamically adjust the number of virtual machines (VMs) to deal with the demand variation. We also simplify the model to only one level utilizing scalarization. The simulation shows the proposed model performs effectively compared to centralized solutions.