Micro-Service Provisioning in the Multi-Tier SDN-Fog Architecture for IoT Applications

Abstract

As the need for fast processing and reply increases, the requirement for high response system increases. It is expected that the internet of things (IoT) devices will be more than two times in number as they are now in a time span of less than 20 years, which makes it harder to deliver services with the same speed existing now rather than faster. Fog Computing provides a distributed platform to data processing and resource provisioning for IoT services, and ensuring quality of services (QoS) in terms of energy and low-latency for these services is essential taking into consideration that deploying IoT services in the Fog computing infrastructure is a difficult process due to heterogeneous fog nodes (FNs). The rapid development of the IoT applications makes human life comfortable, but the IoT applications such as Health-care, Smart Home, Smart Transportation, and Industry have different needs concerning latency, real-time response, makespan, computing capability, etc. Recently, proposed Fog-to-Cloud multi-tier computing model aims to enhance the execution of IoT applications through harmonizing management of underlying resources. In this research, we are proposing a new architecture that can minimize the total makespan of the scheduling algorithms. The four-tire architecture 148proposed in this chapter, i.e., (Cloud, SDN-Controller, Fog-Controller, Fog) might consume relatively more energy, but we aim to prove that it can do the scheduling tasks in less amount of time and less servicing miss ratio compared to the existing architectures, i.e., (Fog-Cloud). In continuation of our work, we propose a services provisioning approach to reduce makespan for IoT applications having service dependency in a multi-tier fog computing architecture.