Estimating Redevelopment Costs for High-Risk Components Using a Genetic Algorithm

Efficient component selection is crucial in component-based software engineering. This selection involves optimally combining commercially available off-the-shelf (COTS) components and in-house components. COTS components are available in different varieties, including low-risk and high-risk components. Among the two, the latter requires significant modifications to be compatible with the system, which can impact productivity, cost, and execution time. This paper addresses the challenge of integrating commercially available off-the-shelf (COTS) and in-house components, with a particular focus on the cost implications of incorporating high-risk COTS components that necessitate significant adaptation. A genetic algorithm-based method is presented to calculate the cost coefficient for redeveloping these high-risk components, following an initial identification based on their reliability. The efficacy of this approach is supported by case study results, which reveal notable improvements in Intra-modular Coupling Density (ICD) and overall system functionality, highlighting its potential to optimize component selection.