The Campus Network Environment Visualization Intrusion Detection System And Implementation

The campus network is essential information construction of school basis hardware facilities. But, safe problem about by link up Internet has brought extremely large challenge to our campus network administrator. Many hacker tools and system bug being exposed unceasingly make our campus network is in danger all the time. Thus, to secure campus network becomes fairly important.This dissertation introduced the campus network environment the visual intrusion detection model,framework and structure, and to achieve a data packet capture module, data extraction module, events analysis database module, user behavior modeling module, visual modeling module, anomaly detection module, the data demonstrate module, security response and visual intrusion detection query modules.This dissertation discuss based on scattered point visual intrusion detection algorithms:the introduction of principal component analysis and curvilinear component analysis as a mapping component analysis tools, a new high-fidelity and deterministic visualization algorithm in intrusion detection is proposed.The principal component analysis mapping as a result of the initial use of iterative algorithm and the use of pseudo-random selection algorithm to increase the point of certainty; mapping and spatial distance to the midpoint between the European Space standardization of the original conversation between the European square from the same goals as the optimization algorithm to improve fidelity.Finally, KDD Cup99 data sets is used to test and certify,results show that the improved curvilinear component analysis algorithm can be used to identify the internal and external intrusion infiltration. The system can be data analysis and process, detection and display of illegal data and storage of suspicious data into log file,the system satisfies intrusion detection system of requirements.