Critical Routers Detection Based On Distribution Similarity Transfer

In infrastructure networks, there are always some critical infrastructures that are in key positions and have large flow. The performance and reliability of the critical infrastructures directly influence the local abilities of the whole networks. However, the critical infrastructures of infrastrcutre networks are often difficult to detect. This thesis aims to develop an efficient approach for detecting critical infrastructures in infrastructures networks. The key contributions of this thesis are as follows.(1) A novel data acquisition system is proposed for acquising network behavior data in enviroments where there are many distributed servers. The architecture of this system is hybrid P2P. In this system, an independent server acts as the console (i.e., deploying configurations and monitoring the tasks). All the other servers involved in the task constitute a P2P network. Considering the influence of routing load balance when measuring the network topology, a Paris-traceroute method is adopted to design the module of detecting routes of this system. In order to meet the requirements of different conditions of network and Internet data center firewall, several different protocols are designed.(2) To improve the ability and security-level of infrastructure networks, we propose a novel method for the critical infrastructures detection problem, which is mainly based on distribution similarity transfer. In real applications, due to several factors (e.g. network status, performance of routers), the behaviors of different routers within different routes often belong to different distributions. Therefore, the proposed method models the problem as the distribution similarity transfer among different routes.First, the suspected routers whose features are different from those of ordinary routers are detected in the target domain (current route) by spectral clustering. Second, a novel distribution similarity transfer classifier is proposed for classifying the suspected routers obtained in the first step. The proposed method is evaluated on the real dataset provided by Huawei Inst. The experimental results have validated the proposed method can effectively detect the critical infrastructures. Meanwhile, it has been demonstrated that the proposed method can successfully adopt the distribution similarity transfer to improve the classification results.