The algorithm design of anti-attack performance testing software in large network

Abstract

This paper studied anti-attack performance testing methods in large network. Anti-attack performance testing process in large networks is different from traditional testing process, which mainly features harbinger attacks, lacking precise characteristic information of determining the attack act. Point to point structure limit characteristics connections, traditional detection method focuses on fixed feature information directly linked to take anti-attack performance testing, once lost contact feature will cause the detection inaccuracy. In order to avoid these shortcomings, this paper proposed an anti-attack performance testing method in large networks based on fuzzy Cmeans clustering algorithm. Collected relevant data to extract and analysis sample characteristics, the use of fuzzy C-means clustering method for classification of data for further calculations, to gain abnormal behavior pattern data to complete anti-attack performance testing in large network. Experimental results showed that using the proposed algorithm for anti-attack performance testing in large network could greatly improve the accuracy of detection, so as to effectively maintain a large network security, and to provide users with good network environment. Introduction With the constantly updated of network technology and range expanding of network applications, network security issues have received more and more attention . Large-scale network is a mainstream network form. Assuming a large network security threats, the network user information and network resources will be destroyed . Therefore, the detection method of antiattack performance in large-scale network has become a mainstream needed research method in network . At this stage, the main anti-attack performance detection method in large networks include the anti-attack performance detection algorithm based on principal component analysis of large networks, anti-attack performance testing method based on quantum neural network algorithm in large networks and anti-attack performance detection method based on ant colony algorithm Large network . Among them, the most commonly used method is based on the quantum neural network algorithm method in large networks . Due to anti-attack performance testing method in large network has a very broad space for development. Therefore, it is able to get attentions of many scholars, and have become hotspot research . Anti-attack performance testing process for large networks is different from traditional inspection process, which mainly features harbinger attacks, lacking precise characteristic information of determining attack act, point to point structure limit the contact between features. Traditional detection method focuses on contact between stationary feature information for anti-attack performance testing, once lost contact feature will cause the detection inaccuracy. In order to avoid the above defects of traditional algorithms, we propose a method for anti-attack performance testing in large network based on fuzzy C-means clustering algorithm. Collected relevant data of sample characteristics to extract and analysis, using fuzzy C-means clustering method for classification of data for further calculations, to gain the behavior patterns of attack data to complete anti-attack performance testing in large network. Experimental results show that using the proposed algorithm for anti-attack performance testing in large network can greatly improve the accuracy of detection, so as to effectively maintain a large network security, and to provide users with good network environment. International Conference on Automation, Mechanical Control and Computational Engineering (AMCCE 2015) © 2015. The authors Published by Atlantis Press 1177 Principle of anti-attack performance detection in large networks Using quantum neural network algorithm can detect anti-attack performance in large-scale networks, to provide users with good network environment. The steps are as follows: Set in a large network, the size of all user data including network operations can be used to describe by 500 r = , the population at large networks anti-attack during performance testing iterative processing times can be used to describe the maximum value max 100 I = . These data were cross-user network operation processing and mutation processing, can obtain an initial population of antiattack performance testing in a large network. For large-scale network operation data crossover and mutation processing, you can get a large population of new network operating characteristic composition. In a large network of anti-attack performance testing, the set of all anomalies can be formed 1 1 1 {( , , ( )), ,( , , ( ))} n n n U z a v z z a w z =  , where, q l z T ∈ , ( ) { 1,1} l w z ∈ − , ( ) 1 l w z w ≤ ≤ , w is running state of a large network, ( ) l x z is the membership of the major network operating characteristics, l ζ is the coefficient variation of a network feature. According to the following formula, it can describe for large networks operating characteristic classification problem: