INTELLIGENT INFORMATION TECHNOLOGY OF DETECTION NETWORK ATTACKS IN REAL TIME SYSTEM

M.P. Komar

The structure and algorithms of neural network immune system of the detection computer attacks are developed, that consists of such subsystem as: traffic pretreatment, training and selection of detectors, traffic analysis and detection of computer attacks, adaptation. The adaptation method of developed system to the trends of computer attacks that can detect new, previously unknown cyber attacks, analyze them and adapt to new realities in order to improve the quality of detection. The implementation of the basic blocks of the neural network immune system of the detection computer attacks is proposed that is based on the modular principle. The experiments on testing the proposed system are made, which show the ability of neural network immune detectors effectively detect polytypic cyber attacks.

Key words: intelligent information technology, network attacks, neural network, immune systems, network traffic, immune neural network detector.