Fog Computing for Energy-Efficient Data Offloading of IoT Applications in Industrial Sensor Networks
Abhishek Hazra, Mainak Adhikari, Tarachand Amgoth, Satish Narayana Srirama
Abstract
Fog computing has recently emerged to <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">in-situ</i> processing and energy-aware data offloading of Internet of Things (IoT) applications in the industrial sensor networks. Besides that, increasing the performance of large-scale IoT applications by improving the emergency response time has become a critical issue in sensor networks. To address the above-mentioned challenges, in this paper, we design a novel Energy-aware Data Offloading ( <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">EaDO</monospace> ) technique to minimize the energy consumption and latency in the industrial environment. The proposed <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">EaDO</monospace> strategy first outlines the emergency information of the incoming tasks with the attribute values. Next, the <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">EaDO</monospace> strategy schedules the emergency tasks using a multilevel feedback queuing policy to improve the schedulability. Moreover, a graph-theoretic approach, called as <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Hall’s</i> theorem is also adopted for finding maximum matching between scheduled tasks and active computing devices, including distributed fog devices and centralized cloud servers. Extensive simulation results exhibit that the <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">EaDO</monospace> strategy significantly improves the energy consumption rate of the industry generated tasks up to 23%-30% over the existing algorithms.