EB-BFT: An elastic batched BFT consensus protocol in blockchain

Abstract

With the rapid increasing number of applications using blockchain, how to improve the throughput of business and reduce the latency of committing transaction on the blockchain have become important topics. This article proposes an elastic batched byzantine fault tolerance consensus protocol whose name is EB-BFT, which elastically changing consensus nodes participating in the consensus and scaling the number of transactions batching in the consensus based on volume of the transaction as well as the physical resources. The protocol EB-BFT first identifies low-performance nodes and suspends them from participating in consensus, so the number of consensus nodes is dynamically changed. After validate the legality of transactions sent by clients, the algorithm of dynamically scaling the number of batched transactions in consensus is proposed. Finally, a series of experiments are conducted to verify the effectiveness and efficiency of EB-BFT in terms of throughput and latency, and it proves when the upper limit of the latency of committing transactions is set at 15 s, the dealing speed of business can reach 20000 transactions per seconds, this results proves that this protocol will make the blockchain having better performance in larger throughput and shorter latency of processing transactions, which also proves that our method can handle transactions as well as traditional centralized system.