Redundancy management of onboard equipment: an arbitration approach based on configuration supervisors

This paper considers the operational reconfiguration of an onboard equipment complex with redundant heterogeneous and non-universal components for achieving fault tolerance and other operational and technical characteristics. An approach to building a redundancy management system (RMS) is formulated as a conceptual solution. This approach uses configuration supervisors according to the number of previously developed competitive configurations of the complex. Each supervisor can be self-sufficient to execute the following functions: monitor the readiness and functional efficiency indicators of the components included in the corresponding configuration; participate in arbitration for the right to implement the corresponding configuration in current conditions; initiate and control the operation of the corresponding configuration. A three-stage algorithm of the RMS is proposed. It contains a sequence of paired arbitrations of computers and configurations. An illustrative example explains RMS operation in different modes of the complex under detectable and undetectable failures of equipment components and computers allocated for redundancy management. The proposed approach unifies and rationalizes the information and logical structure of redundancy management tools, thereby simplifying the creation of an efficient RMS with flexibility, a wide coverage of equipment and configurations, and a significant reduction of bottlenecks in the redundant complex.