Design And Implementation Of Network Security Management System Based On K-Means Algorithm

With the rapid development and improvement of Internet,big data,cloud computing and other technologies,people have entered the era of "Internet +",effectively improving the convenience of people's work,life and study,but "Internet +" is also facing massive attacks and threats,such as blackmail,while providing convenient services for people.Trojan horse,DDOS attacks,etc.,for the application of the Internet has brought great obstacles,and even cause serious property losses.The traditional network security defense technology adopts the active mode,can not actively detect the network latent virus,Trojan horse and other attack threats,only after the outbreak of these attack threats can start anti-virus tools,once these attack threats are highly developed,it will bring serious losses.Therefore,this paper proposes to introduce data mining technology into network security management,construct a network security identification model using K-Means algorithm,discover the potential attack genes in the network in real time and actively,start antivirus software in time,and further enhance the Internet defense capability.The main work of this paper is:(1)the data mining algorithm is studied,and the K-Means algorithm is deeply simulated and analyzed.In this paper,we study the initial centroid selection of K-Means algorithm,propose using density parameters to determine the initial centroid,and on the network security data set,the performance of the improved K-Means algorithm is simulated and analyzed.The experimental results show that K-Means algorithm can accurately detect the potential viruses or trees in the network packets.Horses,such as blackmail,DDOS attacks,Trojan horses,etc.(2)network security management system based on K-Means algorithm.Through interviewing and investigating network security administrators,this paper uses structured modeling technology to derive system logic business functions,which are data acquisition function,data analysis function and post-processing function,then designs the system architecture using three-tier architecture,describes the system data business process,and gives the K-mea.NS algorithm application key process,at the same time for the system database design,provide support for system data processing.(3)implementation of network security management system.In this paper,the data mining algorithm is embedded in the network security management system.Based on MyEclipse integrated development environment,the security management system is developed with Java programming language.The system test shows that it can satisfy the system's good interaction and accurate processing ability,at the same time satisfy the performance requirements of high throughput,high response and high concurrency,improve the real-time,timeliness and reliability of network security management including detecting and killing virus or Trojan horse,and ensure the security of network software and hardware resources.