Search academic texts

Information × Registration Number 0218U000357, 0116U002499 , R & D reports Title Methods for protection against computer attacks based on neural networks, artificial immune systems popup.stage_title Head Sachenko Anatoliy Oleksijovych, Registration Date 26-01-2018 Organization Ternopol National Economic University popup.description2 Research object - information processing process during the classification and detection of computer attacks in information and telecommunication networks. Research subject - protection methods against computer attacks in information and telecommunication networks on the basis of artificial neural and immune networks, fuzzy logic and modular arithmetic. Research aim is the development of methodological and technical foundations of security system against computer attacks based on artificial neural and immune networks, fuzzy logic and modular correction codes to increase reliability of computer attacks detection and classification and improve the level of security. Research methods are system analysis methods, mathematical statistics, application of pattern recognition theory, artificial neural network theory, neuro calculations and artificial immune systems, fuzzy logic, modular arithmetic. Research results: an analysis of known protection methods against computer attacks; a modified method for constructing a detector for detecting computer attacks based on neural networks and artificial immune systems; a method for reducing the dimensionality of information based on deep neural network and using multichannel neural network detectors to construct a hierarchical classifier of computer attacks; a hierarchical classifier of computer attacks was constructed on the basis of combining neural network immune detectors and eliminating conflicts in these detectors operation, which made it possible to determine the class and type of computer attacks; a method of parallelizing the neural network was improved by splitting the training sample into sub-samples and learning each subsample on a separate copy of the neural network model; a parallelization algorithm and a structure of neural network was developed, and experimental studies were carried out that showed an increase in the reliability of attack detection in comparison to known approaches; a generalized architecture of neural network immune system against computer attacks was developed; algorithms and software of the neural network immune system against computer attacks was developed; an experimental study of the developed system was conducted, which confirmed the increase in reliability of detection and classification of computer attacks; an approach to improve the security of the protection system against computer attacks by implementing neural network detectors on programmable logic integrated circuits and the introduction of a decision making subsystem based on Mamdani's fuzzy inference rules was proposed; a method of decision-making in operation of the attacks detecting system based on fuzzy logic; Correction codes based on modular arithmetic were developed that can effectively correct multiple errors in data packets and increase the reliability and security of data transmission in information telecommunication networks; a method was improved and an algorithm for ensuring the integrity of information packets based on modular correction codes was developed. The obtained results can be used to build intelligent security systems against computer attacks that will be able to function and develop themselves independently, without an active participation of the user. Such systems are designed to be used both for protection of personal computers and for protection of networks at various organizations and structures, and are capable of increasing the reliability of computer attack detection by functioning as an additional heuristic analyzer of information in computer network communication channels along with standard signature systems and anti-virus systems or as a separate module within such systems. Product Description popup.authors Івасєв С.В. Биковий П.Є. Головко В.А. Дорош В.І. Дубчак Л.О. Загородня Д.І. Зачко О.Б. Карачка А.Ф. Комар М.П. Кочан В.В. Лендюк Т.В. Марценюк Є.О. Сапожник Г.В. Саченко О.А. Скумін Т.Ф. Турченко І.В. Цаволик Т.Г. Шилінська І.Ф. Якобчук П.Ю. Яцків В.В. popup.nrat_date 2020-04-02 Close