Research On Visualization Of Network Topology Under The Constraints Of Geographic Location Information

With the development of network technology, there are more and more large-scale network topology such as global Internet topology which have tens of thousands of nodes and edges. Traditional text representation approaches of the network topology has been difficult for network managers and network analyzer to understand these network in a very short period of time, but the network topology visualization could show the network topology to the users in the way of image, becoming a powerful tool of network analysis in a short period to time.However, almost all existing network topology visualization methods focus on the structure of the network topology and neglect the actual situation of the network topology, such as network topology node geographic location information. With the development of navigation and positioning system, network manager pay more and more attention to network topology node real information. The visualization of network topology focuses on the abstract structure of the network, and the location information would reflect the real situation of an abstract node. If display the real network topology where ignore the location information, the visualization would show a lot of information of the connection between the nodes and a little real attribution of the nodes which affect the network manager to not know the real network topology fully. In this thesis, there are two main tasks to visualize network topology.1. This thesis present a method based on network topology zoning of geographic information. Firstly, use the community discovery methods to divide the large network topology into some relatively small communities. Each community based on geographic information is arranged in the respective geographical area. This thesis uses knapsack algorithm to layout the communities of no location.2. This thesis presents a layout algorithm aiming at nodes in irregular region. The method uses the attractive operation of the contour nodes and edges to move the nodes in the irregular region, so that these nodes can be uniformly and beautifully arranged in these areas. Because this method takes into account into attractive rather than the outline of the contour shape, the algorithm can deals with the problem of the node layout in random contour shape. Finally, the visualization in this thesis use the above two methods and test the system functional testing. The results showed that, when users need to visualize network topology nodes related to geographic information, the two methods herein combined to guarantee the nodes can clearly show the main structure of the network topology in the network visualization geographical constraints.