EDGE4ALL: edge computing for smart city

Abstract

This paper describes an Edge Computing based architecture for a complex scenario such as a Smart City. A Smart City can be de scribed (from a technical point of view) as thousands or even millions of small disposable devices (IoTs), each one capturing a physical aspect of the environment and endowed with wireless communication capability. Altogether, the generated data will feed intelligent algorithms allowing smart applications, typically running at the Cloud level. Since the Cloud has bandwidth limitations when trying to deal with such a number of connections, it becomes evident the need to introduce an intermediate stage - the Edge - between the devices and the Cloud. Furthermore, in a Smart City context, it is frequent to highlight information security issues that need to be addressed. So, this project aims to develop an Edge, which main requirements are: enforce IoTs security; be solid and robust enough to handle a huge number of logs/connections; efficiently transform low-level data into high-level information; offer a proper API to integrate with higher-level systems; and be easily scalable. Besides describing the design, this paper also presents some results supporting the decisions taken, and discuss some alternative solutions.