Artificial Neural Fuzzy Inference Rule-Based (ANFIS) Model for Offloading Tasks for Edge, Cloud, and UAVs Environment

Abstract

Edge computing is increasingly gaining acceptance in both industrial applications and academic research due to its potential to reduce delays and provide clients with enhanced service quality. However, mobile edge computing presents several significant challenges, particularly given its diverse applications, high user mobility, and the dynamic nature of numerous Internet of Things (IoT) devices. Managing multiple tasks efficiently becomes difficult, especially with the limited resources of these devices. This study introduces a task-offloading model that utilizes an Artificial Neural Fuzzy Inference System (ANFIS) within a layered architecture to address these challenges. The proposed model is designed to optimize task offloading in such dynamic environments. The ANFIS-based model is simulated using MATLAB’s Neuro Fuzzy Logic Toolbox. The layered architecture of the ANFIS model underwent extensive testing, achieving 70% training and 30% testing, and demonstrating an overall low error rate of 0.29505, thereby validating the model’s effectiveness in managing task offloading in mobile edge computing.

Keywords

• Edge
• local
• cloud
• ANFIS
• MATLAB
• task offloading