4. Runtime Validation Framework

Testing large-scale complex systems at runtime is of paramount importance. This is particularly true for dynamical systems, such as distributed adaptive network embedded systems (ANES), which exhibit adaptive capabilities aiming at autonomously reconfiguring and adjusting their behavior based on the changing environmental conditions. In such cases, it is not feasible, during the development stages, to anticipate all the possible operating conditions that the system may face in a real environment. This is because some information about the execution context and the system itself can be available only once the system has been deployed. Thus, in order to correctly assess the effectiveness, efficiency and robustness of ANES, it is required to verify that the system correctly adopts the proper adaptation mechanisms in response to the context changes as well as to check the quality of such adaptations. The focus of the chapter is to discuss about the needs for employing runtime verification and validation of ANES and the main challenges and requirements for its implementation. In addition, it presents a reference framework that supports developers in testing adaptive systems at runtime. One of its key feature is the capability to emulate certain realistic conditions through synthetic data, which is useful to check the system’s behavior under specific and controlled situations.