An Improved PBFT Consensus Algorithm Based on the Raft Voting Mechanism and DAG Ledger Structure

To address the issues of high communication overhead, low throughput, and arbitrary primary node selection in the traditional PBFT consensus algorithm, this paper proposes an improved PBFT consensus algorithm based on the Raft voting mechanism and the DAG ledger structure. By introducing a two-layer architecture composed of proxy nodes and candidate nodes, the system nodes are reorganized. Within each proxy domain, a proxy primary node is elected using the Raft voting mechanism, thereby enhancing the stability and efficiency of primary node transitions. During the consensus process, leveraging the DAG ledger structure enables parallel block generation, which is divided into two phases: data block consensus and address block consensus. Digital signatures and hash commitment mechanisms are introduced in each phase to ensure message integrity and the verifiability of consensus data. Simulation results demonstrate that the improved algorithm achieves lower transaction latency and higher throughput compared to the original PBFT algorithm.