A Study of Detection Probabilities and Real-World Testing of a Human Immunity Inspired Intrusion Detection System

Abstract

In the paper, a study of detection probabilities and experimental tests of an intrusion detection system (IDS) utilizing the artificial immune system (AIS) based on the negative selection approach are presented. The algorithm uses binary strings called receptors to detect intrusions in programs in the operating system. In the work, a statistical approach is used to calculate the probabilities of detecting the intrusions and the results are compared with the experimental tests. New experiments are conducted using a virtual machine as the environment to check the effectiveness of the proposed IDS. For this purpose, real-world malware samples are launched in the virtual machine. The research of the IDS is then presented, analyzed and concluded.