The rapidly changing profiles of traffic on Internet networks require visualization and statistical analysis tools amenable to simultaneous visual contrasts of numerous network designs and performance. Network administration and security management pose major challenges. How should effective strategies be developed for extracting, visualizing, analyzing, and interpreting dynamic Internet traffic behaviors?

Pohl et al. present an integrated visualization and analysis tool (IVAT) for comparing the configuration behaviors of time-varying networks that are poles apart. They investigate algorithms for constructing views of dynamic networks, represented as ordered series of static networks. They implement dynamic network layout algorithms that: compute dragging forces amid connected nodes and propelling forces among every node pair to dispense static network nodes for visualization; uniformly tweak the partition of nodes on dissimilar horizontal network layers; and portray edges as series of upright and straight line segments, to provide better global layout views of static network graphs. The IVAT offers features for plotting, investigating connectivity, semantically transforming, and statistical analysis of different snapshots of dynamic networks.

More practical and sophisticated techniques for data mining and structural modeling of behavior profiles of the Internet backbone traffic exist in the literature [1]. However, the IVAT is still a valuable tool for educating network designers and administrators.