Using Genetic Engineering Algorithm to Prevent Subversion of an Intrusion Detection System

Abstract

Intruders mercilessly attack commercial, academic, defence; healthy centers distributed (networked) systems frequently, and often successfully. The challenges of intruders have become very critical. The most perceived effective defence today is the use of intrusion detection systems. (IDS), though it is widely considered to be impossible to build an effective distributed systems that completely eliminate unauthorized intrusions. It may be effective to thicken the wall of defence by building effective procedures in form of efficient algorithms inside IDSs. The target system should account for misuse detection and anomaly detection by reporting on the presence of an intruder, protecting the system from harm by the intruder, make intrusions into the system difficult, help locate the intruder for a possible prosecution with the law. Our solution analyse each string that is introduced into the computers to those residing in the system already by carrying out a pattern matching with detectors that match only strange patterns and recognise all friendly patterns that are legitimate to the system.