Algorithms Based On A Single Pattern Matching Intrusion Detection System

With the growing popularity of the network, network attacks,network attacks, unauthorizedaccess and computer virus attacks has made the rapid growth of the network security situationbecome increasingly severe, network security and reliability become increasingly concernedabout the problem. And the resulting network security mechanisms, such as anti-virus software,ifrewall, intrusion prevention, intrusion detection, encryption, authentication, security controls.ltis also an important line of defense to protect the network from attacks.However, due to thepresence of two important reasons, traditional intrusion detection systems is diiffcult to assumethis important responsibility.First, as Gigabit,10G optical ifbers appear, the increase in thenetwork bandwidth, the gradual increase in the network traiffc, to complete the real-time iflteringof the data stream by using traditional intrusion detection system has become impossible.Secondly,with the rapid gross of the business use the invasion of the Internet at the same timebecome more diverse, intrusion detection rule base is also getting bigger.So complete rules matchhas become the important challenges of the traditional intrusion detection systems and intrusionprevention systems.This paper introduces the research background and signiifcance of the intrusion detectiontechnology, intrusion detection technology research status, description of the single-patternmatching algorithm research status and intrusion detection systems.The single-mode matchingalgorithm has been focuses research, including BF algorithm, KMP algorithm, BM algorithm,BMH algorithm, BMHS algorithm. It also detailed analysis and presentation for the variousalgorithms. To focus on the inadequacies of the BMHS algorithm, combined with the advantagesof the the BM algorithm and BMHS algorithm we improved the BMHS algorithm and proposed anew algorithm—FSBMH algorithm. The algorithm also use movement of the suiffx jump tableand bad characters jump table to achieve the pattern string move, good suiffx jump tablecombines the matched sting and the last characters of the current window to determine themoving distance of the pattern string good suiffx case, The bad characters jump table can movethe strings to the right up to a distance of2m+l in the best case by using the window sectionsPrinciple pattern. So we can improve the matching eiffciency. We ifnd that the improvedalgorithm is better than the BMHS algorithm in the time complexity and the number of matchesand also improve the eiffciency of intrusion detection by to analysis the improved algorithm byexperiment and compare the experiment results of the improved algorithm with the experimentresults of the BMHS algorithm.