A Novel Multilayer Artificial Immune System for Network Defense [ READ ]
Mohamed M. K. Elhaj, Hussam Hamrawi, Yahia Abdalla
Artificial Immune System is a promising computational intelligence system inspired by human immune system which acts as a natural re-sistance to diseases using sophisticated mechanisms intended to protect our bodies from invaders which are facing a number of layers of defense, i.e. physical, physiological, innate and adaptive layers. Recently we proposed a multilayer network defense artificial immune system inspired by innate immunity in humans, the outcome of that work showed encouraging results from the innate layer. This paper describes a framework of a multilayered network defense system composed of two main layers, innate and adaptive layers, both layers are described, the innate layer as a first layer of defense which is designed and implemented using fuzzy logic expert system, and adaptive layer as a second layer and shows detailed results from the whole system. The innate layer shows very encouraging results since it has already dealt with 77% of the whole traffic with false positive rate of only 0.0107, this rate could still be reduced when adding more rules to the fuzzy knowledge base.
Mohamed M. K. Elhaj, Hussam Hamrawi, Yahia Abdalla
Artificial Immune System is a promising computational intelligence system inspired by human immune system which acts as a natural re-sistance to diseases using sophisticated mechanisms intended to protect our bodies from invaders which are facing a number of layers of defense, i.e. physical, physiological, innate and adaptive layers. Recently we proposed a multilayer network defense artificial immune system inspired by innate immunity in humans, the outcome of that work showed encouraging results from the innate layer. This paper describes a framework of a multilayered network defense system composed of two main layers, innate and adaptive layers, both layers are described, the innate layer as a first layer of defense which is designed and implemented using fuzzy logic expert system, and adaptive layer as a second layer and shows detailed results from the whole system. The innate layer shows very encouraging results since it has already dealt with 77% of the whole traffic with false positive rate of only 0.0107, this rate could still be reduced when adding more rules to the fuzzy knowledge base.
