Intelligent agent based artificial immune system for computer security—a review

Abstract

Since its introduction in the 1990s the internet has proliferated in the life of human kind in many numbers of ways. The two by-products of the internet are intelligent agents and intrusions which are far away from each other in the intention of their creation while similar in their characteristics. With automated code roaming the network intruding the users on one side as worms, viruses, and Trojans and autonomous agents tending to help the users on the other side, the internet has given great research challenges to the computer scientists. The greatest challenge of the internet is intrusion, which has increased and never decreased. There are various security systems for the internet. As the Human Immune System protects human body from external attacks, these security systems tend to protect the internet from intruders. Thus the internet security systems are comparable with human immune systems in which autonomous cells move throughout the body to protect it while learning to tackle new threats and keeping them in their memory for the future. These properties are comparable with that of autonomous agents in the internet. Thus intelligent agent technology combined with ideas from human immune system is a great area of research which is still in its developing phase. In this paper, state of the art of security systems which use both these technologies of intelligent agents and artificial immune system i.e., Agent Based Artificial Immune System (ABAIS) for security are reviewed, paying special attention to features of human immune system used in the system, the role of the agents in the ABAIS and the security mechanisms provided against intrusions.

A Cooperative Negative Selection Algorithm for Anomaly Detection

“…Apart from attractive qualities of AIS such as self-configuration, self-learning, self-adaptation and distributed coordinating, its ability to identify self and non self fascinates research fraternity the most. Lot of research work in AIS is centered around three theories: negative selection, clonal selection and immune theory [14]. New theories are developed and introduced to address different problems of various domains.…”

“…Artificial Immune System (AIS) is inspired from Biological Immune System (BIS) [7,10]. AIS models the processes and procedures of BIS which helps an organism survive under various challenging circumstances [14]. AIS is recognized as a new evolving paradigm for solution of complex computational problems.…”

“…Reaction and response of BIS against new and complex threats motivates researchers and scientists to model anomaly detection after BIS. AIS models various procedures of BIS, out of those Negative Selection Algorithm (NSA) and Clonal Selection Algorithm are the prominent ones [14]. Negative selection is based on the procedure that goes inside the thymus to select T-cells.…”

“…These new clones are not exactly same but they undergo mutation to identify various new pathogens. Afterwards it becomes memory cells and geared up to respond to same pathogen if encountered in future [14]. Negative Selection Algorithm and Clonal Selection Algorithm offer many features that can be exploited to address anomaly detection problem.…”

“…Detectors in NSA aim to cover more and more self and non self space correctly. Detector generation and selection is very important as these detectors forms base for anomaly detection [5,14]. Many previous works focused on it's optimization but still it needs to be addressed properly.…”