SAFER: Crowdsourcing Based Disaster Monitoring System Using Software Defined Fog Computing

摘要

In the occurrence of natural or artificial disasters, conveying immediate and precise situation awareness (SA) information from the salient disaster scenes to the first responders is vital. Crowdsourcing can offer enormous heterogeneous data collected from diverse sources such as sensors, smartphones, vehicles, IoT devices, buildings, and social media. Sending the crowdsourced SA data to the Cloud for analyzation is futile as the opportunity to act immediately may be lost due to delay constraints. However, analyzation of the data-intensive and media-rich contents impose accessibility to massive computational resources, which may be unresponsive in the disaster zone. Fog computing allows quick analyzation of the heterogeneous SA data on the site closer to the disaster scene. This article describes an architecture named SAFER (SDN Assisted Fog computing for Emergency Resilience) which provisions Fog computing and Software-Defined Networking for efficient disaster management. We evaluated the SAFER architecture using simulation tools consuming heterogeneous data and verified that better Quality of Service is achieved by reducing the Service Delay in both transmission and computation. Using SAFER architecture disasters can be detected early than conventional Cloud-based disaster management schemes so that more human lives can be saved.