Generation Algorithm Of Artificial Immune System-based Detector

With the development of computer and technologies, computersystem has been developed to a complicated and interconnectedopening system, which results in more serious problems of intrusiondetection. Intrusion detection system (IDS) is a system thatcontinuously monitors some dynamic behavioral characteristics ofnetwork or computer system to determine if an intrusion has occurred.The intrusion detection technique is an important method to insurethe computer network security. It is becoming one of hot researchtopics in information security field. The operating mechanism ofintrusion detection systems is naturally similar to the human immunesystem. The theory that the immune system can protect body frominvasion provides an important approach to investigating theintrusion detection technique.In recent years, immune-based intrusion detection technology hasbecome a key research field in intrusion detection system. Itsprominent character is that it can explore natural immune logicaltheories, mechanisms and principles for detecting and reacting tointrusions.At present, the majority of commercialized intrusion detectionproducts just adopt the simple template match technique, which canbe only adapted for some simpler attack modes with a highmisrepresentation rate and just can detect known attack modes.Whereas immune-based intrusion detection system can detect unknownattack modes according to immature information, which has greatactual significance.The generation of detector is an important stage in the intrusiondetection system. On the basis of studying the immune theory andthe working mechanism of artificial immune system completely the paper analyses some detector generating algorithms and the papermainly researched on the detection generation algorithms applyingin the process of negative selection. By the analyses on the existingdetector generating algorithms, a new algorithm of detectorgeneration based on Hamming distance is introduced, The efficiencyof detecting is improved by using model and eliminating thereluctant detectors. At last, some experiments of new algorithm aregiven. The results show that the new algorithm can cover the non-selfspace as much as possible and has better affect than traditionalnegative selection algorithm, The work can provide a new method tostudy intrusion detection system. The proposed detector generatingalgorithms has great potential.