Automated measurement-based device traffic modeling in control networks

Traffic models are very important for the performance evaluation of communications systems and decisive for the result quality. The usage of off-the-shelf devices in control networks permits also to use detailed, reusable device traffic models in performance prediction of the network design. However, this requires a high diversification of the device traffic models to support the high number of device types, which can only be realized by automating the model generation. This paper presents a suited approach that automatically abstracts generic device traffic models from reference measurements. Therefore, it efficiently analyzes the device in its sampling behavior and interaction assuming light utilized channels.