The Intrusion Detection Algorithm Based On Artificial Immune Research

Artificial Immune System-AIS is firstly developed for solving the problems in network security. Its many features such as distributed detection, self-adaption, diversity and the memory capacity referred to the Biological Immune System. Negative selection algorithm is very important in AIS that referred to the negative selection principle. It with clonal selection algorithm and network immune algorithm are the core algorithms in the AIS. Currently, negative selection algorithm is widely used for pattern recognition, virus detection, network intrusion detection and some other specific applications.This paper mainly studied the Forrest-negative selection algorithm. To improve the defects that have lager mature-detector set and "holes" in traditional algorithm, a new matching rule was raised. This paper also analyzed the parameters and the effectiveness in this new algorithm.The main work of this paper is summarized as follows:1. Detailed described and made some experiments to prove the impact of all parameters in the negative selection algorithm. The matching rule in the traditional algorithm should be improved.2. Based on the defects in the exisiting algorithm, an r-adjustable threshold fuzzy matching negative algorithm is brought forward. This algorithm has effective detector set and it can eliminate the redundancy phenomen existed in original detector set. At the same time, through adjusting the matching thresholds continuously, it can decrease the number of hols obviously.3. Analyzed the improved algorithm and made some experiments. The results showed that improved algorithm has a good effect..