The Research Of As Level Topology Identification Method Combined With BGP And Traceroute Information

With the rapid development of the Internet network, network has become one of the most important infrastructure in today’s society, and it has become an inseparable part in people’s daily life. Autonomous systems(AS) level of the Internet topology represents the network logical connections between each autonomous systems, network management, network optimization research data base, and it is also the basic premise of a safe network environment. The existing network of autonomous system level topology identification rely mainly on the method through BGP routing information collection, by extracting the AS path attributes of BGP routing table information to get the connection of ASes, to obtain the corresponding grade AS network topology. But in practice, in order to obtain full global AS level network topology, the BGP information is unable to identify many of the network connection of the lower location. With international scholars putting forward using IRR database data to supplement the BGP routing information AS topology data, the disadvantage of the method is the IRR database update too slow that the data are not real-time data, it can not meet the needs of global real-time network topology. In this paper, according to the above problem, this thesis proposes a new AS level network topology identification method, the specific content includes the following three aspects:1) Putting forward a method about the AS level topology combined with BGP information and traceroute information. BGP information is about the control plane of AS level topology, but traceroute information can reflect the actual data of the data flow,it help to obtained the AS connection in the lower place of the Internet.This method can help to get a relatively complete the AS level network topology.2) Adopting a method based on maximization traceroute-BGP path matching: using the data obtained from the traceroute information level is IP network topology data, in order to get the AS level network topology data, this thesis adopted a method based on maximum path matching, according to the source of BGP routing tables AS attribute information, minimize the error rate of the matching to map IP addresses to the corresponding AS, building the corresponding level AS network topology.3) Putting forward an algorithm based on connection relations of combining LG algorithm and DPP algorithm of the AS inference methods: due to the AS level topology identification, it is another way to understand the connection relationships of AS level topology, this thesis combined with the advantages of the two existing algorithms to infer the connection relationship of AS level topology, the accuracy of this method in the inference AS connections has improved to a certain extent.In this thesis, I used the IRL data AS as the original information of BGP routing information of the AS level topology, and the proposed topology identification method, the data show that BGP information and traceroute information combined with the AS level topology identification method can get a more complete AS level topology. Then using the CAIDA theory published data and the proposed AS connection relationships, comparing the inference method show that the proposed method can get more accurate AS connection relationship.