SankatEdge: A distributed network edge infrastructure framework for disaster recovery

Abstract

This paper describes a robust, distributed, virtualized fault-tolerant framework for rapid response to catastrophic events such as earthquakes, land-slides, flooding, fire, and epidemics. The approach is distinguished in that fault tolerance in computing, communications, and storage, and the availability of relevant information content at the edge specific to the disaster scenario and location is ensured. The core themes of the framework are the prepositioning of a rich body of pertinent knowledge to facilitate helpers in a disaster area, and the rapid establishment of ad-hoc wireless networks to ensure that high levels of communication and coordination prevails in the response team, and the affected population. The paper suggests the availability of dedicated servers, local servers, cloud servers, and dynamically deployable servers, capable of support disaster edge services, with a fault-tolerant infrastructure capable of supporting and sharing computing, communications, and storage resources across the network. Virtualization of disaster edge services can be support on such servers if desired. Disaster edge services can be made available in a dormant mode of operation ready to respond to a catastrophe when needed. Adaptation of the proposed system to emerging Network Functions Virtualization (NFV) scenarios is described as well.