Development of Mathematical Models of Immune Networks Intended for Information Security Assurance

Abstract

Abstract : The natural immune system is considered by many specialists as a second brain of vertebrates. In fact, the immune system possesses all the main features of Artificial Intelligence (AI) systems: (1) memory, (2) learning capability, (3) capability to recognize self and non-self, and (4) decision-making capability, that is, the immune system must decide how to treat all macromolecules it encounters even if such molecules are foreign and have never existed before. Of special interest to computer science is the theory of immune networks which describes interactions between immune system specific proteins (antibodies) and foreign macromolecules (antigens). The existence of such immune networks has been established experimentally by molecular immunology which has detected and described the antibody-antigen interaction. Based on the biological principles of the immune system, the field of Artificial Immune Systems (AISs) has been established. It hopes to offer powerful and robust information processing capabilities for solving complex problems. For example, AISs may provide improved techniques to detect and mitigate modern computer network vulnerabilities to intrusions from computer viruses, unauthorized access or other forms of data corruption. Like other modern computer science techniques such as Artificial Neural Networks (ANNs) or Intelligent Agents, AISs can learn new information, recall previously learned information, and perform pattern recognition in a highly decentralized fashion. However, AISs based on natural immune networks differ remarkably from ANNs, intelligent agents, genetic algorithms, and cellular automata in their ability to recognize self and non-self and their highly specific activity. AISs have already been applied to several specific problems including information security, fault detection, robotic control, and others.