Adaptive routing protocol for large-scale power internet of things based on edge computing and federated learning

Area of Science:

• Computer Science
• Electrical Engineering
• Network Engineering

Background:

• Power Internet of Things (PIoT) networks face routing and energy management challenges due to numerous heterogeneous devices.
• Existing solutions struggle with scalability and efficiency in large-scale PIoT deployments.

Purpose of the Study:

• To propose a novel adaptive routing protocol integrating edge computing and federated learning for PIoT.
• To optimize energy efficiency and device coordination in heterogeneous PIoT environments.
• To enhance network throughput and reduce end-to-end delay.

Main Methods:

• Developed a hierarchical architecture with edge nodes for distributed processing and local decision-making.
• Implemented an adaptive routing algorithm adjusting parameters based on real-time network conditions, device specifics, and energy constraints.
• Utilized federated learning for collaborative optimization without centralized data sharing.

Main Results:

• Achieved 35-50% higher network throughput compared to traditional protocols.
• Reduced end-to-end delay by 40-60%.
• Demonstrated 45-65% energy savings.

Conclusions:

• The proposed protocol shows superior scalability and robustness, maintaining performance up to 10,000 devices.
• Edge computing and federated learning effectively address PIoT complexities.
• Significant improvements in network performance and energy efficiency were validated.