Effective disaster recovery for edge computing against large-scale natural disasters

Abstract

With the spread of the Internet of Things (IoT), edge computing (EC) has attracted renewed attention. EC places resources such as CPU and storage closer to the user than cloud computing (CC) to realize low-latency services and/or mitigate network congestion to a cloud. In a telecommunications carrier, resources are placed at telephone central offices (TCOs). At the same time, EC requires disaster recovery for large-scale natural disasters such as earthquakes and tsunami, as TCOs have been damaged by natural disasters in the past. In disaster recovery of CC, data stored in a cloud is backed up to another geographically separated cloud location to prevent data loss due to natural disasters. However, as video IoT which is one of the most popular use cases of EC contains a large amount of data, backing up all data to a cloud causes network congestion to the cloud. Therefore, one way to mitigate network congestion is to back up data utilizing edges (i.e., distributed TCOs). In this paper, we propose disaster recovery for data backup by selecting a data backup location from as many distributed TCOs as possible to mitigate network congestion in EC against large-scale natural disasters. Our proposed method efficiently utilizes resources to prevent data loss due to natural disasters.