Development of crowd monitoring and informing system using fog computing model

Abstract

The concept of future hyper-connected city will include ubiquitous, frequent connections among people, objects, and our environment. In this sense, crowd information is re-interpreted as the level of interactions in which people communicate with physical devices at rooms and stores, and even on streets in various ways. Therefore, it becomes one of the most meaningful and dynamic data that our future data analysis technology will consume. In order to measure and inform crowd information to people, we propose to use an emerging fog computing model, instead of a conventional cloud. A fog moves portions of computing capability of the cloud down to the network edge and form a local computing environment. By distributing computing and networking services closer to where user data is generated, the fog best meet the emerging demands. We develop a testbed of fog computing architecture and implement a crowd monitoring service on top of it. Then, we deploy all the testbed on a public street and run experiments in a real world environment.