A network visualization framework based on multi-resolution animated cone trees

There are many problems to be resolved for network visualization, among which are the topology representation, visualization of data flow between every two directly connected nodes, and visualization of gross data flowing into or out of a node. This paper presents a new network visualization framework based on Multi-Resolution Animated Cone Trees, which can integrate all the three types of information mentioned above into one visualization. First, in the framework, each group of nodes that directly connect to the same upper router node are represented as a cone, and with an effective algorithm for 3D layout of the cones a cone tree is constructed. And then, data flows between nodes are represented as animated colored lines, and the shade of each line's color is determined by the data rates through the line. At last, gross data flowing into or out of a node is represented as a pair of cylinders, whose height represents the quantities of the gross data. The resolution of each cone is determined by the distance from the viewpoint to the center of the cone's bottom, and with flexible interactive methods, focus + context views are constructed. The campus network of Peking University is visualized as an example to verify the efficiency and flexibility of the framework.