Guidelines for Identifying the Next Centralization Step in Multicomputer Architectures

The paper poses a scientific problem regarding the development of multi-computer systems that would be the basis for their use in the field of cyber security and information protection. One of the problematic tasks that needed to be solved was the development of a method for determining the next option for centralization in systems without user intervention in order to complicate the search for the center of the system for attackers and establish the principles of their functioning. As a result of the research, methods for synthesizing systems and systems that are designed to function in corporate networks and can change their architecture during operation, that is, are adaptive, were analyzed. According to the results of the study, insufficient detailing of the internal architecture of systems was established in terms of mechanisms that launch and implement the restructuring of systems, including the center of systems. In the analyzed works, attention is mainly focused on the migration of the center between system components. The choice of the next option for the center of systems is not detailed. Therefore, the task was set in the context of the development of the theory of distributed systems to develop a method for determining the next option for centralization in systems. The work formalized the components and elements of the systems, the connections between them, the operating environment of the systems and their centers, and based on them, rules were developed for selecting the next centralization option. The obtained rules became the basis of the developed method for determining the next centralization option in systems during their restructuring without the involvement of an administrator. A feature of the developed method is the avoidance of complete or significant partial search when selecting a centralization option. To confirm the effectiveness of the proposed solution, an experimental system was developed and a study of centralization options was conducted with it. Also, machine modeling of such a system was carried out. The obtained theoretical and experimental results showed their convergence and confirmed the feasibility of using the developed method. The directions of further research are the development of a systems controller for selecting one and approving the solution options developed in the centers of the systems