ENET: Efficient Blockchain Consensus Algorithm Based on Evil Node Elimination Tree

Abstract

Blockchain consensus algorithms are generally used to ensure data consistency in the blockchain. The current blockchain consensus algorithm has some problems, such as high computing resource consumption, slow block generation speed and difficulty in dealing with malicious node and failure node, which lead to the uncertainty of authenticity of new blocks and trading transaction delay. In view of these problems, we firstly turn the question into solving the problems of electing the one node to keep ledger while ensuring the correctness of ledger information and then synchronize the distributed ledger among most nodes. we propose a consensus algorithm named ENET (Evil Node Elimination Tree) using single chain blockchain structure, with integrating Boyer-Moore majority vote algorithm ensuring the correctness of all new generated blocks. A bottom-up promotion approach to election was adopted to gradually eliminate evil nodes, meanwhile the structure of elimination tree of evil nodes is being constructed. In addition, the distributed synchronization mechanism of top-down update is adopted to recursively update the distributed ledger of all nodes. Theoretical derivation and computer multithreading simulation results show that the communication complexity of this algorithm is at a linear level, and the fault tolerance ratio is 1/4 at the lowest and 3/4 at the highest. The transaction duration is significantly shortened while the double payment problem is eliminated in the system.