The Internet Network Topology Probing Method Based on Classified Strategy

This paper analyzes current network topology probing methods, points out their limitations. Aimed to construct a reasonable large scale network topology probing project, discusses a series of real problems, such as target IP address choosing, redundant data reducing and system frame building. We choose the raw BGP data as the source of probing addresses. Based on the route aggregation, a heuristic classified probing method is proposed to reduce the redundant data in probing process. In order to increase the proportion of network links between different sources probing results, a mutual probing frame is designed to construct circular probing mechanism. In experiment, the validity and superiority of our method in data redundancy is verified. We give a part of visualized result at last.