Proof-of-Activity Consensus Algorithm Based on K-Medoids Clustering

Dong Wang,Chenguang Jin,Bingbing Xiao,Zheng Li,Xin He

DOI: https://doi.org/10.1016/j.bdr.2021.100266

IF: 3.3

2021-11-01

Big Data Research

Abstract:K-medoids cluster-based Proof of Action consensus algorithm (KPoA) is proposed by us in order to mainly decrease the computing power wast and Malicious information dissemination in Proof of Activity consensus algorithm (PoA). In PoA, offline participant nodes cause block headers to be discarded, then the terrible wastage of computing power takes place. Moreover, the efficiency of consensus is greatly affected because the malicious nodes are not handled in time. KPoA uses K-medoids clustering and follow-the-satoshi mechanism to select participant nodes and accounting nodes successively to ensure the unpredictability of accounting nodes. All nodes are grouped by K-medoids clustering, which help to achieve transaction blocks' hierarchical consensus and reduce the spread of malicious information. Multiple accounting nodes take turns to create transaction blocks in the same block header, which nicely increases the number of transaction blocks and reduces the waste of computing power. In order to improve the activity of nodes, the reward and punishment scheme based on credit is set up. The credit value is used to adjust the probability of a node creating blocks. Also malicious nodes can be eliminated in time according to credit value. Experimental results show that KPoA creates transaction blocks at an acceleration of about 2.5 times that of PoA. And its stability is better than PoA.

computer science, information systems, artificial intelligence, theory & methods