The dissertation addresses the new current tasks of creating, maintaining and evaluating the computer systems' security.
The work consists of an introduction, four sections, conclusions, a list of sources used and three applications. The work presented relates to the field of computer systems warranty theory.
The first section, "Developing an Attribution Model for Computer Systems Warranty," discusses the current problems of creating a computer-based computer security system and concludes that a comprehensive approach is needed to address them on the platform of a new fault-tolerant structure. The section provides an in-depth analysis of the main characteristics (attributes) of a CS, namely: uptime, availability, serviceability, survivability, reliability, functional security, confidentiality, integrity with a detailed description of their metrics and methods for quantifying them. Based on the attributive model of the CS (AMGCS), using the metric approach, a methodology for quantifying the level of implementation of attributes, metrics and criteria for evaluating the level of implementation has been developed. In order to analytically evaluate the security level of computer systems, in general, a mathematical model in the form of functional type is proposed.
By considering the attribution model of the guarantee as a set of metric indicators, a methodology for the comparative assessment of computer systems by the achieved level of guarantee of different variants of their implementation has been developed. The basic methods of guaranteeing are described.
The second section, “Designing and Analyzing a Faulty Computer System with a Quasi-Content Structure,” first describes the approach to developing a fault-tolerant computer system for the case where the system created is a critical infrastructure and has the potential for any single failure to suspend execution. its function (in whole or in part) without going into a critical state. Thus, complete fault tolerance is transformed into partial fault tolerance, namely fault tolerance, which enables the design of more efficient systems of critical application.
Basic software requirements for fault-tolerant computer systems have been developed and the ability to achieve a high degree of security for management programs, software bugs and hardware failures has been demonstrated through the use of error prevention, detection, and fault tolerance techniques.
The classification of two-channel fault-tolerant structures of computer systems has been improved, the advantages and disadvantages of each of the structures considered with the logical function of the restoring body "I" are shown. In order to improve the operational readiness of the existing structures, a new quasi-self-repairing two-channel structure was developed with reconfiguration of component parts, consisting of equally reliable duplicate units with restoring bodies having the "AND / OR" function.
It is proved that by decomposition of functional blocks into equally reliable parts and addition of control and reconfiguration circuits it is possible to increase the overall probability of trouble-free operation of the system as a whole. The quasi-content structure easily scales in the direction of increasing the number of duplicate nodes and has a high potential for use in the construction of trouble-free systems of high availability.
In the field of creation of computer systems with high survivability, the application of the principle of infrastructure redundancy and topological optimization is demonstrated, using the example of a quasi-capacitive structure in the construction of a hydroelectric power station automatic safety system. A high survivability cluster structure has been developed, which has significant prospects for the development of endurance critical cluster systems.
In the third section, "Ensuring the failover of fault-tolerant computer systems", we investigated the reliability and reliability of a Quasi-Bridge Structure (QBS), and its advantages over other dual-channel structures. This is confirmed by analytical calculations by several methods and by statistical modeling. The reliability of the QBS functioning was evaluated using the original phenomenological model, which demonstrated its advantages over other two-channel redundant structures.
Section 4, Microprocessor Control and Information System for Blagovest Railroad Crossings, discusses the problems of accident at railroad crossings, and addresses their solutions through the introduction of a new model-based control and information system that enhances traffic through railroad crossings.
Within the framework of competition of scientific and technical projects of scientific institutions of NAS of Ukraine in IPMMS of NAS of Ukraine.
Keywords: dependability, fault-tolerance, fail safety, attributive model of dependability, quasi-bridge structure, quantitative assessment of the level of dependability.
