Font Size: a A A

Study On Intrusion Detection System Based On Clustering And Neural Network

Posted on:2010-01-18Degree:MasterType:Thesis
Country:ChinaCandidate:Q WangFull Text:PDF
GTID:2178360302459371Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
With the increasingly wide application of computer and network technology, especially transmit of governmental and military information in the network; the network security problems are more and more remarkable. Because new attacks ceaselessly emerge with the intrusion technologies, firewall and other passive security methods cannot provide complete protection. As an important and active security mechanism, Intrusion Detection reinforce the traditional system security mechaniam. Intrusion detection techniques can help us detect attacks by monitoring the behavior of users, networks, and computer system. By monitoring and analyzing, abnormal and illegal activities can be discovered, including attack using security vulnerabilities by legitimate users or unauthorized access. In addition, Intrusion Detection System (IDS) can diagnose which type of attack or malicious activity is taking and response to stop the intrusion.Most current intrusion detection system with single-level structure can only detect either misuse or anomaly attacks. Some IDSs with multi-level structure or multi-classifier are proposed to detect both attacks, but they are limited in adaptive learning. In this paper, two hierarchical IDS frameworks clustering and neural network are adopted to detect misuse and anomaly attacks.Firstly, this paper reviewed the status and common network security technology, analyzed status and development of Intrusion Detection in detail, and introduced the basic theory, features and shortcomings existed in them.Secondly, this paper analyzed means of intrusion, described its principle and methods in detail. This paper emphasized the necessity of intrusion detection, introduced methods of anomaly and misuse detection, analyzed structure of intrusion detection system.Thirdly, the paper proposed an anomaly detection model based on clustering. This paper described the principles of ant colony algorithm,then, the paper introduced implementation and simulation of the ant colony optimized algorithm; contrast to the one calculated by BP.Finally, this paper proposed a model based on cubic spline weight function neural network. By analyzing the advantage of cubic spline weight function neural network, designed a misuse detection model. This method solved the shortcomings of BP network.
Keywords/Search Tags:Network Security, Intrusion Detection, Intrusion Detection System, Anomaly Detection, Misuse Detection, Cubic Spline Weight Functions Neural Network
PDF Full Text Request
Related items