RMG-PBFT: A Consensus Mechanism Based on Reputation and Random Matching for Medical Device Traceability

In response to the high communication complexity and latency issues present in current Byzantine fault-tolerant algorithms applied in the field of medical device traceability, this paper proposes an improved RMG-PBFT consensus algorithm based on reputation value grouping and probabilistic matching. Firstly, a node reputation assessment model is developed to enable dynamic grouping of nodes during the consensus process, thereby enhancing consensus efficiency. A reputation recovery mechanism is introduced to ensure that the system can effectively recover in the face of sporadic errors or fluctuations in node performance, thus improving fault tolerance. The three-phase consensus process of the PBFT algorithm is optimized through the implementation of a dual-node communication mechanism based on probabilistic matching embedded in the Commit phase, which effectively reduces communication redundancy and resource consumption, while decreasing the likelihood of collusion among malicious nodes. Finally, some improvements to the view change protocol in the consensus are also presented. Experimental results indicate that the proposed RMG-PBFT consensus algorithm demonstrates better performance in terms of latency and throughput. This method shows favorable results in reducing communication overhead and delays among nodes, as well as mitigating the influence of malicious nodes within the traceability process.