Towards the Evaluation of the Performance Efficiency of Fog Computing Applications

Abstract

Nowadays, Cloud Computing represents an effective solution for organizations that need to adopt a model of elastic computing services and pay-as-you-go. However, in new domains, such as Cyber-Physical Systems and the Internet of Things (IoT), which manage large volumes of data, the time of latency in communication between the Cloud server and users, as well as data processing costs, can be a limitation. Therefore, it is proposed the Fog Computing architecture, which is used as a mediator among IoT devices that generate and demand data (e.g., cellphones, video cameras, sensors) and Cloud infrastructure. The purpose of Fog Computing is to relieve the Cloud server load and to delegate it to intermediate devices. In this context, it is essential to measure the performance to evaluate the solution and exploit all the advantages provided by Fog nodes. This paper presents a performance efficiency quality model for Fog Computing applications. The model breaks down in sub-characteristics and attributes with defined metrics and thresholds that should be used to evaluate these kinds of applications to understand how efficient is the use of computational resources (i.e., time, processing, memory). Also, it is offered an evaluation method that is supported by the quality model to assess the software development process for Fog Computing applications. The evaluation method is aligned with the ISO/IEC 25040 standard and indicates the developer steps towards a good quality Fog Computing solution.