Optimizing Internet of Things Fog Computing: Through Lyapunov-Based Long Short-Term Memory Particle Swarm...

Area of Science:

• Computer Science
• Network Engineering
• Artificial Intelligence

Background:

• Internet of Things (IoT) systems face challenges with high latency in edge-to-cloud communication.
• Increasingly, IoT devices are both data producers and consumers, demanding efficient resource management.
• Traditional cloud computing models struggle with the demands of real-time smart services.

Purpose of the Study:

• To introduce a novel layered computing network based on fog computing principles.
• To develop an algorithm for optimizing user tasks and allocating computing resources in rechargeable networks.
• To address performance limitations and optimize the utilization of low-power edge devices.

Main Methods:

• A hybrid algorithm combining Lyapunov-based control, dynamic Long Short-Term Memory (LSTM) networks, and Particle Swarm Optimization (PSO).
• Fog servers dynamically train LSTM networks for predictive task data feature forecasting.
• Lyapunov functions are utilized for dynamic resource control in rechargeable networks.

Main Results:

• The proposed algorithm demonstrates superior energy efficiency and resource allocation optimization compared to traditional methods.
• Predictive task allocation based on forecasted data features improves unload decisions.
• Optimized utilization of low-power devices and effective resource management in rechargeable networks.

Conclusions:

• The novel fog computing approach significantly enhances the efficiency and user experience of IoT systems.
• The algorithm effectively reduces latency and energy consumption, overcoming edge device limitations.
• This research contributes a robust solution for managing resources and tasks in complex, rechargeable IoT environments.