Leader Election Optimization in Raft Consensus for Private Blockchain Networks

Aim: Leader election is essential for maintaining stability in private blockchain networks. The Raft consensus algorithm, while widely used, faces challenges like frequent leader failures and unnecessary re-elections, which disrupt network performance. This study improves Raft by introducing Activity Score-based leader selection and an adaptive timeout mechanism to enhance reliability and efficiency. Methods: The proposed model ranks candidate nodes based on real-time performance metrics such as transaction throughput, uptime, and response time, ensuring that the most stable node becomes the leader. It also dynamically adjusts election timeouts to prevent unnecessary leader transitions and improve network responsiveness. The approach is implemented in a Hyperledger Fabric-based private blockchain environment and tested against the standard Raft algorithm using key performance metrics. Results: The enhanced model improves leader stability, reduces election frequency, and increases transaction throughput. It boosts network stability by 10%, cuts leader replacements by 28.6%, and raises transaction throughput by up to 5%, ensuring more reliable blockchain operations. Conclusion: The enhanced Raft consensus model provides a more stable and efficient approach to leader election in private blockchain networks. By reducing unnecessary re-elections and optimizing leader selection, the model ensures higher availability and improved transaction processing. These results demonstrate its potential for real-world applications in finance, supply chain management, and healthcare. Future research should explore further scalability enhancements and integration with hybrid blockchain frameworks.