The Reputation Model Assesses Node Validators Consensus on the Federated Byzantine Agreement Blockchain using Eigen Trust
Erdhi Widyarto Nugroho, Mustafid Mustafid, Luhur Bayuaji · 2024
The method for validating blockchain consensus has progressed from the initial Proof of Work (PoW) system implemented by Bitcoin to the Byzantine Fault Tolerance (BFT) family. At present, the Federated Byzantine Agreement (FBA) is under development due to its superior speed, higher throughput, and greater openness in comparison to PoW and Practical Byzantine Fault Tolerance (PBFT). FBA primarily relies on the Quorum Slice, or a cluster of validator nodes. The trust among the validator nodes forms the basis for this group’s formation. A validator node’s degree of reputation directly correlates with its level of trust among the public. Stellarbeat.io provides an index that serves as a benchmark for evaluating the reputation level. This index is generated by computing the mathematical mean of six factors. One problem with this reputation index is that its findings lack fairness due to the presence of validator nodes that lack a quorum set but still achieve a high index score. Furthermore, the lack of diversity in the index values poses challenges in establishing rankings. The objective of this paper is to create a reputation evaluation model for FBA validator nodes by EigenTrust as a viable option for calculating the reputation index and addressing the index problems at stellarbeat.io. The methodology in this paper entails gathering data from the stellarbeat.io API over a period of three months, mapping the validator nodes according to the quorum set, and executing the EigenTrust procedure. The result is a reputation evaluation model for FBA validator nodes using EigenTrust, which provides a more equitable and straightforward ranking of reputation assessment outcomes compared to the stellarbeat.io index. .