Scholarlink

<p>As a peer-to-peer digital payment system, Bitcoin leverages anonymity to protect user privacy. However, beyond this original purpose, anonymity also obscures the real information of a malicious person, facilitating illegal activities such as money laundering, ransomware, blackmail, and darknet. From the perspective of a supervisor, it’s imperative to trace and analyze addresses. This anonymity hinders the traceability of illicit activities, while entity recognition, as a key component of de-anonymization, can enhance the ability to trace and regulate malicious activities. In this paper, we design a scalable complex network-based entity aggregation and recognition system named <b>BiEnTracer</b>. We model and experimentally analyze the Bitcoin transaction network by combining complex network theory. Based on the constructed network, BiEnTracer integrates multiple heuristic algorithms to achieve comprehensive address aggregation and utilizes an address cluster graph to associate clustered addresses with entities, thereby revealing the actual user identity behind transactions. Considering the continuous expansion of the Bitcoin transaction network, we propose a novel incremental entity aggregation algorithm based on label propagation. This algorithm can update entities in linear time, thus improving the scalability of BiEnTracer and enhancing the traceability of illegal activities. Extensive experiments demonstrate that our system exhibits superior entity aggregation performance in complex transaction network scenarios, with an overall aggregation rate reaching 56.4%, and an empirical study is conducted to reveal the transactional behavior characteristics of Bitcoin entities from multiple dimensions, including entity scale, distribution, and activity. The results indicate a highly uneven distribution of entity sizes in the Bitcoin network: 14 large entities control over 100,000 addresses, and some independent addresses exert considerable influence. The majority of entities hold balances below 1 BTC, with assets highly concentrated among a small number of major holders. Furthermore, large-value transactions primarily occur between highly active entities, forming a tightly interconnected network.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

BiEnTracer: a scalable complex network-based entity aggregation and recognition system in bitcoin network

  • 期刊论文

  • 发表日期：

    2025年8月7日

查看全文
  • Li Li,
  • Hao Wang,
  • Wei Cui

摘要

As a peer-to-peer digital payment system, Bitcoin leverages anonymity to protect user privacy. However, beyond this original purpose, anonymity also obscures the real information of a malicious person, facilitating illegal activities such as money laundering, ransomware, blackmail, and darknet. From the perspective of a supervisor, it’s imperative to trace and analyze addresses. This anonymity hinders the traceability of illicit activities, while entity recognition, as a key component of de-anonymization, can enhance the ability to trace and regulate malicious activities. In this paper, we design a scalable complex network-based entity aggregation and recognition system named BiEnTracer. We model and experimentally analyze the Bitcoin transaction network by combining complex network theory. Based on the constructed network, BiEnTracer integrates multiple heuristic algorithms to achieve comprehensive address aggregation and utilizes an address cluster graph to associate clustered addresses with entities, thereby revealing the actual user identity behind transactions. Considering the continuous expansion of the Bitcoin transaction network, we propose a novel incremental entity aggregation algorithm based on label propagation. This algorithm can update entities in linear time, thus improving the scalability of BiEnTracer and enhancing the traceability of illegal activities. Extensive experiments demonstrate that our system exhibits superior entity aggregation performance in complex transaction network scenarios, with an overall aggregation rate reaching 56.4%, and an empirical study is conducted to reveal the transactional behavior characteristics of Bitcoin entities from multiple dimensions, including entity scale, distribution, and activity. The results indicate a highly uneven distribution of entity sizes in the Bitcoin network: 14 large entities control over 100,000 addresses, and some independent addresses exert considerable influence. The majority of entities hold balances below 1 BTC, with assets highly concentrated among a small number of major holders. Furthermore, large-value transactions primarily occur between highly active entities, forming a tightly interconnected network.