A Framework for Optimizing Effort in Testing of System of Systems

Abstract

The concept of System-of-System (SoS) is increasingly used in building large and complex software systems which essentially provide a set of required functionalities in a box to the customers. Testing System-of-Systems includes testing of individual systems, validating the interactions among the constituent systems and verifying the compliance of the service level constraints. Testing complete interactions of constituent systems and checking the compliance with service level constraints as and when they are plugged into the SoS requires significant effort. Therefore, there is a need for testing methodologies to validate such complex systems in a comprehensive and optimal way. In this paper, we propose a test framework based on knowledge characterization of constituent systems that are used to reduce the overall effort in validating the system of systems. The knowledge is characterized in terms of the different test parameters of constituent systems. A theoretical case study has been carried out on a next generation Network File System along with its constituent file systems to illustrate the efficacy of our framework.