ABSTRACT: Today’s Internet provides a globaldata delivery service to millions of end usersand routing protocols play a critical role inthis service. It is important to be able toidentify and diagnose any problemsoccurring in Internet routing. However, theInternet’s sheer size makes this task difficult. One cannot easily extract out themost important or relevant routinginformation from the large amounts of data. collected from multiple routers. To tacklethis problem, we have developed Link-Rank, a tool to visualize Internet routingchanges at the global scale. Link-Rank weighs links in a topological graph by thenumber of routes carried over each link andvisually captures changes in link weights inthe form of a topological graph withadjustable size. Using Link-Rank, network operators can easily observe importantrouting changes from massive amounts of routing data, discover otherwise unnoticedrouting problems, understand the impact of topological events, and infer root causes of observed routing changes.
KeyWords: Network visualization,information visualization, Internet routing,interactive graphics, data analysis, visualmining
KeyWords: Network visualization,information visualization, Internet routing,interactive graphics, data analysis, visualmining
