A Multilayer Graph Model Of The Internet Topology - Dataset (2012)

<strong>Content:</strong> This dataset contains an integrated snapshot of the Internet Topology (year 2012) with corresponding graph models at the IP, Router, PoP, AS and ISP layers. <strong>Purpose:</strong> Despite intensive research during the last two decades, the detailed structural composition of the Internet is still opaque to researchers. Nevertheless, due to the importance of Internet maps for the development of more effective routing algorithms, security mechanisms, and resilience management, more detailed insights are required. This article advances the understanding of the Internet structure by integrating data from different large-scale measurement campaigns into a set of comprehensive Internet graphs at different abstraction levels, and analyzes them in terms of important statistics and graph measures. <strong>Design/methodology/approach:</strong> This study follows the topology measurement framework suggested by Gunes and Sarac (2009), involving three phases: topology collection, topology construction, and topology analysis. <strong>Findings:</strong> An integrated data set of Internet graphs at different abstraction layers is provided that can serve as a baseline for future research on Internet analytics. Furthermore, results of important graph metrics are presented and power-law relationships for the degree distributions on every level of the current Internet are substantiated. <strong>Research limitations/implications:</strong> By necessity, the integrated graphs provide a snapshot of the Internet topology. In future work, repeated measurements and automated data integration could lead to a better understanding of Internet dynamics. <strong>Practical implications:</strong> Due to increasing dependency on the Internet as a critical global infrastructure, studying Internet connectivity is more important than ever for both companies and Internet service providers. The data set will be made publically available for network research. <strong>Social implications:</strong> Understanding the structure of Internet serves as a fundamental step in improving the robustness, security, and privacy of any online service. <strong>Originality/value:</strong> By carefully integrating six different traceroute datasets such as iPlane, CAIDA, Carna, DIMES, RIPE Atlas, and RIPE IPv6L, this paper presents the Internet graphs of a substantially larger and thus solid scale than previously known, at well-established abstraction levels such as the IP interface, router, Point of Presence (PoP), Autonomous System (AS), and Internet Service Provider (ISP). Furthermore, by employing a broad diversity of graph measures, this study creates a more exhaustive snapshot of the global Internet topology than earlier works.<strong>数据集内容：</strong>本数据集包含2012年互联网拓扑（Internet Topology）的整合快照，以及对应于IP层、路由器（Router）层、PoP（Point of Presence）层、自治系统（Autonomous System, AS）层与互联网服务提供商（Internet Service Provider, ISP）层的图模型。 <strong>数据集用途：</strong>尽管近二十年来学界开展了大量相关研究，但互联网的详细结构组成对研究者而言仍不透明。鉴于互联网拓扑图谱对于开发更高效的路由算法、安全机制与韧性管理方案具有重要意义，学界亟需获得更细致的互联网结构认知。本文通过整合多组大规模测量实验所得数据，构建了一系列不同抽象层级的完整互联网图结构，并基于关键统计指标与图论度量对其展开分析，从而加深了学界对互联网结构的理解。 <strong>设计/研究方法：</strong>本研究遵循Gunes与Sarac（2009）提出的拓扑测量框架，包含三个阶段：拓扑数据采集、拓扑结构构建与拓扑分析。 <strong>研究发现：</strong>本研究提供了不同抽象层级的互联网图结构整合数据集，可作为未来互联网分析研究的基准参照。此外，本文还给出了关键图论度量的计算结果，并证实了当前互联网各层级的度分布均符合幂律（power-law）分布关系。 <strong>研究局限与启示：</strong>受限于现有条件，本整合数据集仅为互联网拓扑的某一时刻快照。未来可通过重复测量与自动化数据整合，进一步深化对互联网动态特性的认知。 <strong>实践意义：</strong>随着全球各界对作为关键基础设施的互联网依赖程度日益加深，无论是企业还是互联网服务提供商（ISP），研究互联网连通性的重要性都达到了前所未有的高度。本数据集将面向网络研究领域公开发布。 <strong>社会意义：</strong>明晰互联网结构是提升各类在线服务的鲁棒性、安全性与隐私保护水平的基础前提。 <strong>创新性与价值：</strong>本文通过精心整合iPlane、CAIDA、Carna、DIMES、RIPE Atlas与RIPE IPv6L等6组不同的路由追踪（traceroute）数据集，在IP接口、路由器（Router）、接入点（Point of Presence, PoP）、自治系统（Autonomous System, AS）与互联网服务提供商（Internet Service Provider, ISP）等成熟的抽象层级上，构建了规模远超以往研究的可靠互联网图结构。此外，本研究通过采用丰富多样的图论度量方法，生成了比此前相关工作更为全面的全球互联网拓扑快照。