Checkability of hierarchical transmitions for behavioral check

Abstract

The means of on-line testing of complex distributed information systems (DIS), often possessing criticality properties, must ensure their complete verifica-tion in real time. The high structural-functional and combinatorial complexity of modern DIS translates most of the tasks of their construction and analysis into the NP-hard class. To a large extent, this refers to the design and analysis stages of the system-architectural, structural-functional, and especially behavioral levels. This state is even more aggravated by the dynamism, non-determinism, intelli-gence of situational component interactions of DIS. In designing DIS, the issues of ensuring the reliability of their functioning play a crucial role. Existing and promising models and methods for checking and diagnosing DIS, providing their solution, inherit these features, in addition, they can often exceed the complexity of the corresponding models and design methods. These circumstances necessi-tate the use of decomposition, network and hierarchical models, multi-level, in-cremental and parallel methods that reduce the dimensionality of the tasks of building and analyzing DIS, in particular, behavioral check and diagnosis. This reduction is particularly significant with operational behavioral on-line testing performed in real-time DIS operation. The article discusses certain aspects of de-composition hierarchical models of behavioral on-line testing - formal models of inheritance of testability in the hierarchical transitions of a multi-level behavioral check model of DIS based on hierarchical Petri nets. This model provides a defi-nition of the mechanisms for interlevel inheritance of the identified check behav-ior, which is essential for the hierarchical temporal decomposition of on-line tes- ting. Inheritance is based on the use in hierarchical transitions of compatibility and quasi-order relations for alphabets of positions, transitions, input-output al-phabets and associated conditions, events, actions, functions of hierarchical Petri nets when organizing reference recognition of their behavior.