The improvement of wavefront cellular learning automata for task scheduling in fog computing

Abstract

The rapid advancement of the “Internet of Things” (IoT) devices has led to the emergence of different types of IoT applications that need immediate response and low delay to operate. The emergence of fog computing has provided a proper platform to process fast‐emerging IoT applications. Nevertheless, to name the disadvantages of fog computing devices, it can be said that they are typically distributed, dynamic, and resource‐limited. Therefore, it seems a substantial challenge to schedule fog computational resources effectively to perform heterogeneous and delay‐sensitive IoT tasks. The problem of scheduling tasks aimed at minimizing the energy consumption of fog nodes is formulated in this article, while meeting the requirements of the quality of service (QoS) of IoT tasks, including response time. Minimizing the deadline time and balancing the network load are also considered in the mathematical model. In the next stage, a new algorithm is introduced based on a wavefront cellular learning automata (WCLA) called the wavefront cellular learning automata improved by genetic algorithm (WCLA + GA). WCLA + GA is indeed a modified version of WCLA that has been improved using the genetic algorithm. In this version, the WCLA reinforcement signal is regulated by a genetic algorithm that accelerates the automata convergence rate. WCLA + GA is then utilized to schedule fog tasks. Simulating the proposed method followed by comparing it with other methods demonstrates that WCLA + GA performs task scheduling significantly better in terms of response time, energy consumption, and percentage of tasks that meet their deadline.