An Adaptive Intrusion Detection Algorithm Based on Improved Dynamic Clonal Selection Algorithms

Intrusion detection systems´ adaptability and diversity have been researched for long time. With the development of computer immunology, the dynamic clonal selection algorithm is tried to solve the problem. Based on some improved dynamic clonal selection algorithms, an adaptive intrusion detection algorithm is presented in this paper. According to the algorithm, an intrusion detection system is composed of a self-body antigen set, a memorial immunocyte set, a mature immunocyte set and an immature immunocyte set. An immature immunocyte grows into a mature one if it goes through self-tolerance. A mature immunocyte grows into a memorial one if it matches enough non-self-body antigens in limited time and it goes through co-stimulation. A memorial immunocyte doesn´t die until it can´t go through co-stimulation. Immature immunocytes are generated with clone and hypermutation methods when necessary. The self-body antigen set is renewed during above co-stimulation. Simulation experiments prove that the algorithm have good adaptability and diversity