The goal of this research is to investigate the optimization of the Cell based software architecture. Cell-based software architecture structures a software system into interconnected cells, each comprising multiple elements. This study focuses on optimizing the architecture by determining the optimal number of cells and their internal organization. To achieve this, the Community Detection approach, which identifies closely connected elements, was applied. To preserve cell boundaries, reduce complexity, and enhance modularity, we introduce the concept of functionality, which can be represented by one or more cells. This concept serves as the foundation for optimizing software architecture. A series of experiments were conducted to analyze the problem dimensions that can be addressed through optimization and to evaluate the robustness of the mathematical model. Given that the proposed model is unable to solve large-scale problems efficiently, we developed a heuristic approach and compared its results with those obtained from the mathematical model. The evaluation results indicate that different software architectures can be derived in terms of cell granularity, composition, and interaction. Since each cell can contain multiple elements realized in various architectural styles, the proposed model enables the integration of diverse architectures within a single software system. This flexibility enhances the system?s adaptability and overall efficiency.
Milić et al. (Thu,) studied this question.