Research On The Application Of Artificial Immune Algorithm In Intrusion Detection

Network security problems have brought serious damage to the state,attaches great importance to the network security problems all over the world.Due to the rapid spread of network intrusion,destroyed the social stability,disrupted the normal order of military,political,and allows users to severe losses.Current network intrusion protection relies on the antivirus software,antivirus gateway,etc.Virus software due to the increase in virus feature library,the software will consume more and more computing resources and storage resources,make its detection efficiency are greatly constrained.On invasion of these protective ways are mainly used in network intrusion feature matching technology,to known network intrusion has high detection rate,but unknown to the emergence of new network intrusion detection rate is low,can't solve the problem of changing the network security.With unknown network intrusion in the form of diversification,the number of network intrusion to accelerate growth,research based on the theory of artificial immune adaptive and self-learning of intrusion detection is very necessary.The immune system to recognize its antigen and antigen of the "other",and the antigenic alien microbes or other foreign body produces an immune response,has three important functions: immune from stability,immune surveillance and immune defense,the three function can let the system to remove antigen,maintain its own internal environment consistent goal.Abstract,artificial immune system and the function of network intrusion detection system,network intrusion detection system also have similar three big functions,so it can draw lessons from immunology theory to study and development of intrusion detection system,the reference method can not only improve the defense capability of the system,but also decrease The Times of intrusion detection of false alarms for network management and running open up new avenues.Ideal of intrusion detection system needs to meet the function and requirement: detect network running status,identify and prevent external invasion of network attack behavior,automatically eliminate redundant false-negatives,reduce the network load.The main problems of current network intrusion detection systems:1)The adaptive detection capability is not strong,low degree of intelligence.2)The detection accuracy is poor,omission of(False Negatives)and False Positives(False Positives)rate is higher;The efficiency of the detection system is due to produce a large number of false alarm,behavior recognition is also contributing to the network load.Aiming at these problems this paper proposes a modified artificial immune optimization algorithm,using the improved artificial immune algorithm to reduce the intrusion detection and processing in the process of false alarms generated by the new method,the algorithm is based on negative selection algorithm in,introducing the difference method respectively,local outlier factor fitness fitness,fuzzy rules,such as technology,their advantages to improve the model of artificial immune algorithm,improve the detection accuracy and intelligent degree,optimize the detection index,reduce the false alarm.By simulation and experiment,this algorithm results comparing with the typical NSA classification algorithm,the algorithm proposed in this paper has certain advantages,for the development of network intrusion detection technology provides a useful reference.