We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Statistical and clustering analysis of attributes of Bitcoin backbone nodes.
- Authors
Xu, Dawei; Gao, Jiaqi; Zhu, Liehuang; Gao, Feng; Zhao, Jian
- Abstract
Bitcoin is a decentralized digital cryptocurrency. Its network is a Peer-to-peer(P2P) network consisting of distributed nodes. Some of these nodes are always online and in this article are called Bitcoin backbone nodes. They have a significant impact on the stability and security of the Bitcoin network, so it is meaningful to analyze and discuss them. In this paper, we first continuously collect information about Bitcoin nodes from July 2021 through June 2022 (which is the longest duration of data collection to date). In total, we collect information on 127,613 Bitcoin nodes. At the same time, we conclude that the fluctuation of Bitcoin nodes is directly related to the fluctuation of onion network nodes. Further, we filtered 2694 Bitcoin backbone nodes based on our algorithm. By analyzing the backbone nodes' attributes such as geographic distribution, client version, operator, node function, and abnormal port number, it is demonstrated that these nodes are centralized and play an important role in the Bitcoin network. Based on this, three unsupervised machine learning algorithms are selected to cluster multiple attributes of backbone nodes in a more scientific way. In this paper, the whole process from data collection to cluster analysis is completed and the best results are obtained by comparison. The experiments proved the existence of centralization of Bitcoin backbone nodes and obtained the number of nodes within each cluster. Finally, cluster nodes are de-anonymized based on the optimal results. Through our experiments, we obtain organizational information about the deployers of 103 nodes, linking the Bitcoin backbone nodes to the real world, thus accurately demonstrating the existence of Bitcoin centrality.
- Subjects
BITCOIN; CLUSTER analysis (Statistics); MACHINE learning; STATISTICS; PEER-to-peer architecture (Computer networks); CRYPTOCURRENCIES; COMPUTER network security; SPINE
- Publication
PLoS ONE, 2023, Vol 18, Issue 11, p1
- ISSN
1932-6203
- Publication type
Article
- DOI
10.1371/journal.pone.0292841