The determination of 2-optimal digraphs set for a one-step diagnosis of system

To identify faulty elements during the diagnosis of computer network one can use the diagnostic model so-called PMC model ( Preparata, Metze, and Chien ). The idea of self-diagnosis of a network through the PMC model is based on setting a method of mutual testing of the system's elements, described in the described graph, which we called the graph of diagnostic inference (GDI). GDI, which enable to identify m faulty elements of system, is called an m-diagnosable graph, while the m-diagnosable graph with minimum number of arcs is called an m-optimal digraph. The determination of such digraphs is a NP-complete problem, particulary in reference to the heterogeneous systems, which computer networks usually are. As a rule, a computer network executes various functional tasks on each individual computer. The algorithm to determine a 2-optimal GDI of an order higher then fifth is based on the knowledge of same parameters of the fifth order graphs class. This paper presents an analytical method of determining this GDI class, as well as all 26 digraphs of this class.