Enhancing Byzantine Fault Tolerance in Blockchain Networks through Dynamic Clustering

In recent years, blockchain technology, which enables transactions to be distributed across multiple computers and managed in an immutable and secure manner, has garnered significant attention. Within blockchain networks, consensus mechanisms ensure the consistent sharing of ledger information when new blocks are added. In consortium blockchains, typically employed by a limited number of organizations, the Practical Byzantine Fault Tolerance (PBFT) protocol is widely used. PBFT is designed to tolerate Byzantine nodes-nodes that may be compromised or malfunctioning-by achieving consensus as long as fewer than one-third of the total nodes are Byzantine. However, PBFT relies on the assumption that at least two-thirds of the nodes behave correctly, making consensus challenging when the number of malicious nodes exceeds this threshold. Previous research has explored the use of clustering to enhance throughput, but these methods are static and unsuited to dynamic environments. Moreover, clustering techniques aimed at bolstering Byzantine resistance remain underexplored. This paper presents a novel method for constructing clusters within a blockchain network to resist Byzantine nodes. By employing clustering, we estimate the locations of potential attackers, thereby enhancing the system’s resilience to Byzantine faults.