The Computer Virus Detection Which Based On The Collection Relation Of Classification And Coverage

Computer virus detection has been a hot issue in the field of computer security.Traditional virus detection technology has been unable to meet the demand for real-timedetection of computer viruses.Computer virus detection is similar with artificial immunesystem，simulate the mechanisms of the biological immune system which successed inprotect the body’s success,make the artificial immune principle applied to computer virusdetection,to improve the performance of the virus detection.Since the negative selectionalgorithm was proposed,a large number of scholars has made deep study on virus detectionprinciples and techniques，the area of the virus detection has made considerableprogress.The core technology elements of the negative selection algorithms include thefollowing aspects:the representation of data and detector;the choice of matchingrules;detector generation mechanism;the storage form of the detector.The generationmechanism of detector is the core issue to the virus detection technology.This paper wasdeep analyzed in the principles and technology of the virus immune detection,focusing onthe negative selection algorithm for depth profiling,improved the mechanism of the detectorgeneration and hole detection.Focus on the malpractice of the existing detection mechanisems which exist a lot ofredundant and time consuming much,proposed a high performance detector generatingalgorithm which is based on the collection division and coverage.Based on the r valuesegmentation to extract the self collection of complement,improved the connectionstrategy，as far as possible to achieve the division of the collection，generated the minimumdetectors to cover the largest nonself space，decrease generation time of the detector,reducethe number of detectors，expand the coverage range of equal number of detector，improvedthe detection results.Simulation results demonstrate the efficiency of detection and the timeand space superiority.Any kind of matching rules are inevitably produce detection holewhich has serious impact on the performance of virus detection algorithm.A new methodsis proposed for the detection of hole,this method used the collection information ofcomplement and autologous, this method can detect all the vulnerabilities，the collection ofthe hole is completeness.Established a computer virus detection model based on theimproved virus detection algorithm,and use experiments to verified the performance of themodel.Experimental results show that the virus detection model based on improvedalgorithm for known virus，unknown and known viruses variants have a high detectionresults.