An Analytical Approach to Compute the System Architectural Complexity

Abstract

The management of system complexity is essential to ensure the successful delivery of a system. However, as system complexity continues to grow, an accurate quantitative measure of the complexity of a system becomes a necessity. In this study, we consider the definition of structural complexity introduced by Sinha, and we derive analytical expressions for the complexity of both individual components alone and their interactions. Among the complexity metrics developed to date, we considered entropy-based ones, which allowed us to objectively quantify the complexity while satisfying the criteria introduced by Weyuker. The notion of intrinsic descriptive complexity introduced by Kolmogorov, instead of the Shannon entropy, was used to analyze the structural complexity of engineered systems. The applicability of the structural complexity metric to the graph representing the system architecture is demonstrated.