METHOD OF RESTORING CRITICAL INFORMATION SYSTEM SERVICES ON A MOBILE PLATFORM UNDER CONDITIONS OF CONTROLLED DEGRADATION

The subject of this work is methods for restoring critical services of an information system on a mobile platform under intermittent connectivity and controlled degradation. The objective of the study is to develop a service-oriented recovery method that ensures a measurable reduction in the shortfall in meeting service quality threshold requirements within connectivity windows, taking into account resource constraints and the ability to complete service state reconciliation within the current connectivity window. The study used the following methods: system analysis, mathematical modeling of service states and profiles, formalization of the shortfall in meeting service quality threshold requirements, a predicate description of the window feasibility of recovery steps, and a computational experiment to evaluate recovery performance metrics. The article aimed to solve the following tasks: formalize a recovery step as a transition between service profiles that reduces the shortfall in meeting service quality threshold requirements; introduce a condition for the window feasibility of a recovery step as a combination of resource feasibility and the possibility of completing service state coordination within the current connectivity window; to construct a rule for forming a set of recovery steps within a window that ensures a reduction in the deficit provided at least one step is executed and is consistent with permissible managed degradation actions as a peripheral resource release mechanism. Achieved results consist in the implementation of a window feasibility formalism and a system of metrics for evaluating compliance with feasibility conditions, the proportion of windows with deficit reduction, the cumulative deficit of critical services, and recovery time, which makes it possible to distinguish environmental constraints from the properties of control decisions. Conclusions: Ignoring the possibility of completing service state coordination within the current connectivity window leads to the emergence of incomplete recovery transitions and discrepancies between the expected recovery effect and the actual deficit dynamics, whereas the proposed method identifies the steps taken within the current connectivity window and ensures a reduction in the deficit of critical services in windows where recovery is possible. This enhances survivability by reducing the duration of degradation episodes of critical functions and decreasing the cumulative deficit. The practical value lies in the possibility of using the method for onboard information systems of UAV groups and other mobile nodes with connectivity windows of variable duration and bandwidth.