【Author】 Maeng, Soohoon; Essaid, Meryam; Lee, Changhyun; Park, Sejin; Ju, Hongteak
【Source】INTERNATIONAL JOURNAL OF NETWORK MANAGEMENT
【Abstract】Ethereum is arguably the second most popular cryptocurrency-based network after Bitcoin. Both use the distributed ledger technology known as the blockchain, which is considered secure. However, the provided security level is proportional to the number of connected nodes, the number of influential nodes, and the supported amount of hash power. Thus, the knowledge of the network properties and nodes' behavior is helpful to protect the network from possible attacks such as double-spending attacks, DDoS attacks, 51% attacks, and Sybil attacks. This paper proposes a node discovery mechanism, which performs a P2P link discovery on the Ethereum main network. For that, we develop Search-node, a modified version of Ethereum client that searches for all participating nodes in the blockchain network, stores the node information in the Bucket, and then processes the peer discovery method. Based on the collected data, we first visualize the Ethereum network topology and analyze the attributes of the network such as node degree, path length, diameter, and clustering coefficient. We then analyze the node properties and provide analytical results regarding the relationship between nodes, heavily connected nodes, node geo-distribution, security issues, and possible attacks over the influential nodes. As a result, we have identified 68,406 nodes with a total of 642,034 edges. By analyzing the collected data, we have found that the diameter in the Ethereum network is equal to 8. The node degree is over 19, which is two times higher than the default configuration.
【Keywords】
【标题】以太坊P2P网络拓扑和对等体属性的可视化
【摘要】以太坊可以说是仅次于比特币的第二大加密货币网络。两者都使用被称为区块链的分布式分类账技术,该技术被认为是安全的。但是,提供的安全级别与连接节点的数量、有影响的节点的数量以及支持的哈希能力成正比。因此,了解网络属性和节点行为有助于保护网络免受可能的攻击,如双重花费攻击、DDoS攻击、51%攻击和Sybil攻击。本文提出了一种节点发现机制,该机制在以太坊主网上执行P2P链路发现。为此,我们开发了Search-node,这是一个改进版本的以太坊客户端,它搜索区块链网络中所有参与的节点,将节点信息存储在Bucket中,然后处理对等发现方法。基于收集到的数据,我们首先将以太坊网络拓扑可视化,分析网络的节点度、路径长度、直径、聚类系数等属性。然后,我们分析节点属性,并提供关于节点之间的关系、重连接节点、节点地理分布、安全问题和可能对有影响节点的攻击的分析结果。因此,我们已经识别了68,406个节点,共有642,034条边。通过分析收集到的数据,我们发现以太坊网络的直径等于8。节点度大于19,是默认配置的2倍。
【发表时间】2021
【收录时间】2022-04-23
【文献类型】Article
【论文大主题】链上数据分析
【论文小主题】交易网络可视化及分析
【期刊级别】SCI四区
【影响因子】1.914
【翻译者】王佳鑫
评论