Two Features Encoded In Immunology-based Ids

Artificial Immune has now become the focal point of computation and intelligence in recent years. This paper explores a extensible and universal model based on the artificial immune system . The research of applied foundation in science and technology office of SiChuan which is the subject of pivotal technology in intelligentized intrusion detecting system provides the foundation for this project. This paper describes the differences between two encoding, which are real-encoding and binary-encoding. Then combining this two encodings to build a new detection model which based on data-merging.First, discuss the actuality and direction of information security and intruding detection system before that. Information security is not only a question discussed by academia, but also the question affecting closely people's ordinary life today. In this paper, we also show the significance and austerity of information security. The actuality and direction of intruding detection system is discussed and we will see that the main disadvantage of present-day intruding detection system is lacking for the intelligence and the ability to adapt the changes of network and detect the undiscovered intrusion. So the first question we will resolve is how to make IDS adaptive by studying the changes of network.Biological Immune System can resolve this difficult problem in theory. BIS has so many excellent characteristics such as diversity,stabilization,remembrance,adaptive and so on. These characteristics inspire so many mans who engage in IDS or information security. Recently, the problems in BIS and IDS or information security have amazed comparability. So the application of BIS in information security attracts prevalent attention.The innovation of this paper is importing the idear of data merging which is used to combine real-encoding and binary-encoding correctly and build a mechanism of real-valued negative selection which can use the normal samples as the input to generate the abnormal samples automatically.Finally, at the part of experiment, the process,data,parameter of arithmetic are analysed and a new detection model is studied, which can be patulous as a node in some distributed system in order to improve the ability of detection.