DoSP: A Deadline-Aware Dynamic Service Placement Algorithm for Workflow-Oriented IoT Applications in Fog-Cloud Computing Environments

Abstract

The next generation Internet of Things (IoT) applications are offering multiple services and run in a distributed heterogeneous environment. In such applications, Quality of Service (QoS) requirements are in jeopardy when the computing operations are only outsourced to the public cloud. For IoT applications, a comprehensive framework that supports QoS-aware service placement in a fog computing environment is highly required. It is a challenging task to orchestrate the time critical IoT applications in the fog environment. To alleviate this problem, this paper proposes a novel multitier fog computing architecture called Deadline-oriented Service Placement (DoSP) that provides the services both in fog and cloud nodes. This research work proposed a methodology to utilize low-cost fog resources while ensuring that the response time satisfies a given time constraint. It uses the Genetic Algorithm (GA) to dynamically determine the service placement in the fog environment. In this work, we used the iFogSim simulator to model DoSP and measured the impact of the service placement technique in terms of service deadline. It has been observed that through the proposed solution, there is a reduction in service execution delay, i.e., approximately 10.19% of the overall response time to the EdgeWard and 2.58% to the Cloud Only.