Graph Complexity Measures as Indicators of Verification Complexity

Increasing system complexity affects the complexity of their verification strategies, cognitively overloading engineers throughout the system development life cycle. A graph-based approach has been developed to allow a scalable verification strategy complexity analysis through the use of fundamental patterns. Through verification strategies graphs, this paper proposes the possibility of a mathematical measurement for verification strategies independent from system complexity in terms of requirements and their verification. Two verification strategies on real-world systems were used to showcase the scalability and resource efficiency of the approach. Ordinal comparison between the two graphs revealed that there were mathematical graph complexity measures correlated with the complexity of verification strategies they represent, with their fundamental patterns providing additional information on their differences. These correlations indicate that verification strategy complexity is connected to their graph complexity, validating the feasibility of a quantitative measurement applicable to existing real-world applications and future system developments.