Research And Application Of Novel Artificial Immune Algorithm

Artificial immune algorithm imitates biological immune systems in protecting animals from viruse, bacteria, parasite and dangerous foreign pathogen, which has a lot of advantages including, distributed, robustness, tolerance, adaptability, diversity, immune feedback and self-organization, etc. Artificial Immune System generally constructs based on negative selection algorithm, which generates the mature detector distinguish between self and non-self samples via imitating the T cell development process in the thymus. However, traditional negative selection algorithm has a lot of redundant detectors and holes of detection, and lack self-adaptation, which may cause poor efficiency and accuracy of detection.This paper analyses the factor influencing accuracy of detection and efficiency in details, and presents a cutting based negative selection algorithm in order to reduce redundant detectors and holes of detection, set the threshold value by the characters of different regions, and improve efficiency and accuracy of detector; In addition, we also design the multi-layers management mechanisms of detector to improves the self-adaptation by optimizing the efficiency of searching and updating detectors. At last we realize lightweight access control system using by the new algorithms and establish the security disk prototype system to provides high I/O performance. Evaluating I/O performance, the result proves this novel algorithm has higher efficiency and accuracy of detection. The main contents of this paper are as follows:1) This paper analyzes the factors in detector generating, detector management and matching rules, which influence the efficiency and accuracy of detection about negative selection algorithm.2) We presents the cutting based negative selection algorithm which cuts space of domain by spatial location of self to generate detector, then improves efficiency of detector generating, eliminates redundant detector, reduces hole of detection, and ensures high efficiency and accuracy of detection.3) Designing the dynamic threshold value algorithm by the characters of different regions to optimizes the efficiency and accuracy of detection.4) Designing the multi-layers detector storage structure, the priority-based management mechanism of detectors and the fast detector updating algorithm, which improve the efficiency and self-adaptation of negative selection algorithm.5) In the intelligent disk system we realize the function of access control using by the new algorithms and establish the security disk prototype system. By comparing the I/O performance the result proves this novel algorithm has higher efficiency and accuracy of detection at the same time that it can ensure the security of system.