Measuring the maintainability of Commercial Off-the-Shelf (COTS) based systems: A complexity approach
Smith, Michael W.
MetadataShow full item record
This dissertation presents a predictive approach to evaluate the maintainability of a Commercial Off-the-Shelf (COTS)-based system by analyzing the complexity of the deployment of the system. The approach integrates architectural dependencies and the system’s concept of operations to derive a network-based representation of the software system. A greater understanding of the deployment complexity is gained by using a Design Structure Matrix to determine the number of architectural dependencies on a COTS product, or in-degree, for each COTS product in the system. The arithmetic mean of the in-degree for all nodes in the system can then be compared with the perceived effort to maintain the system. The resultant measure is useful in evaluating the maintainability of the operational system while the system is being designed and throughout its lifetime. Architects can use the approach to assist in COTS product selection and to make product trades to optimize the maintainability of the system. Integrators can use the approach to optimize product deployment and to determine the upgrade strategy for deployment. Finally, maintenance engineers can use the approach to estimate the effort required to maintain the system and to identify areas in which extensive product expertise is required. Because the approach requires only basic information about the system, it can be applied early in the design process and used until the system is decommissioned.