A Distributed Intrusion Detection System Research Based On Biological Immune Mechanism And Mobile Agent

As information technology continues to evolve and network applications continues to increase, information network has become an important infrastructure for social development, involving government, military, cultural and educational fields. Network information security has also become the core which has constrained and influenced the network development, and is more and more important. The threats of computer network security are growing day by day, and insecurity network is constituted by many factors, including human factors, natural factors and accidental factors. Accordingly, the computer user's security awareness and security means to be improved.In recent years, intrusion detection, as a network security technology, has also become a focus of research. Intrusion detection system is automatic and real-time intrusion detection and response system in the computer network. Without interference of the network transmission process, it can provide maximum security for the network. As means of attack becomes increasingly sophisticated and attack technology becomes increasingly mature, intrusion detection techniques have become increasingly demanding. Traditional intrusion detection systems have not solved the current network security issues. To this end, designing a new, intelligent intrusion detection system and trying to make computer networks more secure become the main task of Iinternet information security research.Traditional intrusion detection systems have a number of shortcomings, for instance, weak detection of unknown attacks, a high false alarm rate, the system's single point of failure and the system's own security risks. In this thesis, the principle of biological immune system and the technology of mobile agent are combined to make up for its shortcomings. By using the biological immune character, and adopting both congenital response mechanisms and adaptive response mechanisms, this thesis uses memory detector to improve system checks on the efficiency and accuracy of known invasion. Through the evolution of some memory detector can quickly detect known attacks which pass through simple genetic variation. Through continuous learning and evolution, adaptive detectors have produced the candidate detectors and themselves by negative selection, the system thus has good adaptability, and able to detect unknown attacks. Using the mobility and self-reproduction characteristics of distributed mobile agent system it overcomes the phenomenon of a single point of failure and enhances the security of the system.Through theoretical study and simulation we show the robustness and effectiveness of the model. It can make up for some shortcomings of the traditional intrusion detection system. The results indicate that the model is easy to achieve and has a good performance.