A Monitoring and Prediction Model of Workflow Based Self-Adaptive Software System

It is critical of self-adaptive software system to monitor all kinds of information in real-time and obtain relevant data of the system to make the system can adjust automatically when necessary. In order to cope with the current market needs and rapidly changing environment, and combine the Web services with self-adaptive software systems perfectly, we proposed a monitoring and prediction model for workflow based self-adaptive system (WSAS for short) based on our previous work, which gives users more flexibility in expressing requirements on the model level, provides dynamic QoS value calculation and prediction at run time, and can both monitors operating status of system in real-time and monitor the context of the system, allowing software systems to better meet users´ changing needs, and the operating environment changes. We also proposed ratio dependent and variable exponential smooth prediction method for QoS based on above model. Finally, through some experiment, it confirms that the prediction method of this paper is feasible and effective.