Authors:
Mayssa Trabelsi
and
Samir Ben Ahmed
Affiliation:
LIPSIC Laboratory, Faculty of Sciences of Tunis, University of Tunis ElManar, Tunis, Tunisia
Keyword(s):
Internet of Things (IoT), Fog Computing, Task Scheduling, Real-Time, QoS Optimization.
Abstract:
With the increasing demand for real-time processing for IoT applications, Fog computing becomes a crucial approach to overcome the limitations of centralized Cloud Computing. Given its decentralized structure, Fog computing enables faster response time, real-time processing, and reduced latency, making it particularly suitable for time-sensitive IoT applications. In this paper, we propose a novel approach called the ”Energy-cost-aware task scheduling with a Deadline-constrained” (ECaTSD) algorithm for real-time task scheduling in a fog infrastructure. The main objective of the proposed algorithm is to minimize energy consumption and monetary costs under deadline constraints. The ECaTSD algorithm dynamically allocates incoming tasks to the most suitable fog nodes in real-time. It selects the fog node that meets deadline requirements with the least energy consumption and monetary cost in the infrastructure. Moreover, the proposed algorithm has been simulated using the iFogSim simulator
. The algorithm’s performance is evaluated using various criteria, such as the percentage of IoT tasks successfully meeting deadlines, energy consumption, monetary cost, and response time compared to other scheduling policies. ECaTSD algorithm shows high efficiency in meeting deadlines (99.58% completion rate) while being energy and cost-efficient.
(More)