Efficient data consensus algorithm integrating FL and blockchain dynamic partition protocol PBFT.

Area of Science:

• Computer Science
• Information Security
• Distributed Systems

Background:

• Internet of Things (IoT) data sharing faces challenges like privacy breaches, single points of failure, and high communication overhead.
• Existing federated learning lacks robust collaboration and fault tolerance, while traditional PBFT algorithms struggle with high concurrency and communication complexity.
• There is a critical need for secure and efficient data transmission solutions in the rapidly evolving IoT landscape.

Purpose of the Study:

• To propose a multi-layer data sharing framework for IoT environments that integrates federated learning and blockchain technology.
• To enhance data privacy, security, and transmission efficiency in IoT networks.
• To address the limitations of existing solutions regarding fault tolerance and communication overhead.

Main Methods:

• Implemented a framework combining federated learning with blockchain technology.
• Utilized differential privacy for data privacy and Proof of Quality (PoQ) consensus for fault tolerance.
• Introduced a reputation mechanism for node supervision.
• Developed an improved PBFT consensus algorithm with dynamic region partitioning for efficient data transmission.

Main Results:

• Differential privacy reduced data transmission accuracy by only 3.5% with Gaussian noise.
• The PoQ algorithm demonstrated a 14.3% fault rate for fault-tolerant nodes.
• Optimized PBFT reduced communication frequency by 44% and transaction latency by 63% compared to single-layer PBFT.
• The framework achieved efficient IoT data transmission with real-time security.

Conclusions:

• The proposed multi-layer framework effectively addresses security and efficiency issues in IoT data sharing.
• The integration of federated learning, blockchain, differential privacy, PoQ, and optimized PBFT offers significant advantages for high-concurrency IoT scenarios.
• The solution meets real-time security demands and improves fault tolerance and communication efficiency in IoT networks.