一种基于层次化聚类学习的区块链共识方法

A block chain consensus based on hierarchical clustering learning

为解决目前主流区块链中普遍采用的POW、POS或者PBFT等共识算法存在算力浪费或者TPS处理效率低等问题,本文提供了一种利用层次化聚类学习在区块链中达成共识的方法。该方法具体步骤如下:确定区块链中带聚类节点筛选指标;根据指标,计算每个待聚类节点的特征值;根据每个待聚类节点的特征值,利用机器学习对所述多个待聚类节点进行循环聚类处理,确定出目标节点;目标节点即为区块链共识节点。该方法可降低区块链达成共识的时间,提高共识效率。

To solve the resource waste and low TPS speed problems in block chain, design a new Consensus algorithm based on Hierarchical Clustering Learning. Including： choose the Cluster peers from the block chain ; Determine the eigenvalues of each of the nodes to be clustered.; according to the eigenvalues of each node to be clustered, the duster nodes are clustered to identify the target nodes. According to the throughput and node operation of block chaining, new machine learning can be carried out, and appropriate indexes can be selected according to samples.