An efficient sharding consensus algorithm for consortium chains.

Area of Science:

• Blockchain Technology
• Distributed Systems
• Network Security

Background:

• The Practical Byzantine Fault Tolerance Algorithm (PBFT) faces scalability issues in large consortium chains due to communication congestion and reduced throughput with increased nodes.
• Existing consensus algorithms struggle to balance decentralization, security, and scalability in consortium blockchain environments.

Purpose of the Study:

• To propose a novel clustering-based sharding consensus algorithm (KBFT) for consortium blockchains that enhances decentralization, security, and scalability.
• To address the limitations of traditional algorithms like PBFT in large-scale network environments.

Main Methods:

• Implemented a K-prototype clustering algorithm for node sharding based on mixed attributes.
• Utilized disjoint transactions for parallel consensus within different shards.
• Introduced a supervision and node credit mechanism for enhanced security and proxy node selection.
• Analyzed shard size selection using binomial probability distribution and node failure probabilities.

Main Results:

• The KBFT algorithm demonstrates significant improvements in scalability and throughput compared to PBFT.
• Achieved a marked reduction in communication complexity.
• The proposed algorithm maintains system security and robustness.

Conclusions:

• KBFT offers a more efficient and scalable consensus mechanism for consortium blockchains.
• The integration of clustering, sharding, and credit mechanisms enhances overall system performance and security.
• KBFT provides a viable solution for large-scale consortium chains requiring high throughput and robust security.