Research And Analysis On Network Security Detection System Implementation Based On B/S Architecture

With the improvement of Internet technique in today’s society, Network security has become a widespread issue of concern. Internet is an open platform, it has some characteristics such as Participants diversification, the connection form diversification, these characteristics make the Internet easy to suffer from outside intrusion, and the security is under great threat.With the development of enterprise information, many application systems use the B/S structure, so the research on B/S system architecture based on the unique significance. Aiming at the shortcomings and deficiencies of the network security detection technology nowadays, this thesis introduces data mining technology into the system, using a distributed network security monitoring analysis model developed with B/S architecture. With the help of these technologies, the accuracy, real-time of the Internet Security Detection System enhanced.This thesis firstly introduces the classification, the technology of intrusion detection system structure nowadays, and introduced several major elements of performance evaluation of network security detection system which is the current mainstream. Then this thesis describes the basic principle and process of data mining technology, and introduces in details of the system structure of the mining. Based on these technologies, the data mining algorithm is introduced into the network security detection system; this thesis introduces K-means algorithm and genetic algorithm and its improvement.In the basic theory and algorithm, this thesis designs the network security detection B/S architecture model based on analysis of the system architecture. This thesis describes in detail the design idea of network security detection system and detailed design of the key technology of safety inspection system. Then, this thesis has carried on the design of the main function modules of this system, Including the underlying Snort detection phase, collecting network data standardization stage, improved K-means clustering stage and the improved genetic optimization stage.Finally, this paper carries on the simulation experiment aimed at the various modules of the system. The experimental results show that the system has strong processing ability. Compared with other systems, this system has markedly improved in the accuracy, real time. This system effectively reduces the system false negative rate.