Ensuring Reliable Network Communication and Data Processing in Internet of Things Systems with Prediction-Based...

Area of Science:

• Computer Science
• Network Engineering

Background:

• Distributed Internet of Things (IoT) systems and fog computing necessitate reliable communication for Quality of Service (QoS).
• Failure to meet QoS in real-world interacting IoT systems can lead to critical failures and negative user impacts.

Purpose of the Study:

• To introduce a prediction-based resource allocation method for Multi-Access Edge Computing (MEC)-capable networks.
• To ensure required QoS and optimize resource utilization for diverse IoT use cases, adapting to dynamic user requests.

Main Methods:

• The method utilizes current resource load and historical data to predict future utilization.
• Resource allocation optimization criteria are adjusted based on predicted changes for incoming requests.
• Developed for edge computing scenarios like autonomous vehicle data exchange with predictable demand fluctuations.

Main Results:

• The proposed approach demonstrates increased reliability for processes within IoT systems.
• Effective management of periodic resource demand fluctuations, such as during peak hours, is achieved.

Conclusions:

• The prediction-based resource allocation method enhances the robustness of IoT systems.
• Optimized resource utilization and reliable QoS are achievable in dynamic edge computing environments.