Generation of detectors and detecting method

Abstract

The invention relates to generation of detectors and a detecting method. The method comprises the following steps: defining a maximum interval as an initial detector; then carrying out self-tolerance training on the initial detector; according to matching rules, removing a point matching a known self from the detector, simultaneously dividing the interval of the detector from the point to generate two subinterval detectors and recursively detecting each subinterval. The detectors are optimized in the process and the generated mature detectors are finally gathered to be used for system security protection. The invention breaks through the existing detector generation method and matching rules and eliminates loopholes and redundant detectors, thereby further improving the generation efficiency and detection efficiency of the detectors.

Description

technical field [0001] The invention applies various excellent characteristics of the biological immune system to the computer system, and is mainly used to solve practical problems such as computer virus prevention and control, network intrusion detection, etc., belongs to the cross technical field of bioinformatics and computer science, and specifically relates to a detection generation and detection methods. Background technique [0002] The immune system protects the body from external bacteria, viruses, etc., and is able to recognize foreign cells or molecules, and then eliminate these harmful substances from the body, while observing the appearance of abnormal cells in the body, and eliminating cells that have mutated. The primary problem facing immune cells is how to define self / nonself and recognize them. Self / non-self identification is a high-efficiency pattern recognition problem. [0003] At present, the research on the artificial immune system includes two cate...

AI Technical Summary

Problems solved by technology

[0010] In 2005, He Shen et al. proposed a detector generation method with a variable detector length l. This method eliminated the "hole" area and improved the detection efficiency of the detector, but there was still the problem of redundant detectors. And the generation of its detector is based on the L-level R-tree for search and matching, the required time complexity and space complexity are still high

[0011] To sum up, since the original antigens and antibodies are expressed in the form of binary strings that are difficult to process, the current detector generation methods all have problems of low detector generation efficiency, loopholes and redundant detectors to varying degrees.