Adaptive Immune-Based System For Network Security

Although a lot of effort has been made and a lot of studies about network security have been carried out, the security requirements for different type of network are increasing daily because of the huge amount of malware and hacking mechanisms are developed daily. Features that must be used in network to provide a high level of protection are; detection, protection and recovery. As such, this study include these features and it is inspired from human immune system, by simulating the mechanism of the natural immune system in computer network to provide security and work as adaptive immune system which learn by themselves. in this research we are concentrating more to learn the system by training the system to distinguish between self and oneself pattern by using different mechanism as calculating checksum of the packet and the checksum of system 32 files, and compare the incoming packet with the created database.. The system can observe any change occur in system 32 and trace any modification, creation, deleting and renaming inside system 32 When the system in learning phase the number of false positive are increase, but after the system gains experience, the false positive rate will decrease gradually. The results indicate that the proposed system is very accurate and is capable of tracking any change and detecting any malware within the system as well as high efficiency.