Hierarchical Artificial Immune Model

As viruses become more complex, current anti-virus methods are inefficient to detect various forms of viruses, especially new variants and unknown viruses. This chapter proposes a hierarchical artificial immune model (HAIM) for virus detection to overcome three specific shortcomings in traditional artificial immune system (AIS) models: randomly generating the detectors leading to the bad efficiency; poor generalization and poor performance with a big dataset; and ignoring the relevance between different extracted signatures in one virus. The virus gene library generating module works on the training set consisting of legal and virus programs. The model can obtain the frequency information of deoxyribonucleotides (ODN) appearing in the legal and virus programs. Finally, classification decision is an overall behavior that greatly reduces the information loss. The model can effectively and efficiently recognize obfuscated virus, detect new variants of known virus and some unknown viruses.