Analysis of Internet topology data

Discovering Internet topology is important for analyzing routing protocols and Internet robustness and resilience. Recent research results dealing with Internet topology, such as the discovery of power-laws and the application of normalized Laplacian analysis to Internet topology data, have increased the need for more complete datasets and their more rigorous interpretations. In this paper, we examine datasets from two sources: Route Views and RIPE. We show that each dataset may have a geographical bias and, consequently, consist of distinct routes. Furthermore, by employing normalized Laplacian analysis, we identify distinct cluster characteristics that could not be inferred by directly examining the collected data.