Calibration of deployment simulation models : a multi-paradigm modelling approach
Faculty of Sciences. Mathematics and Computer Science
Faculty of Applied Engineering Sciences
S.l. , 2012
TMS/DEVS '12 : proceedings of the 2012 Symposium on Theory of Modeling and Simulation, Society for Computer Simulation International, San Diego, Calif., USA
University of Antwerp
In embedded systems development, software engineers increasingly rely on modelling and simulation to produce optimal design solutions. A bottleneck in the Modelling and Simulation Based Design (MSBD) process is model calibration. Setting up experiments to estimate parameter values such that the model accurately reflects real-world system structure and behaviour is technically complex and labour intensive. Parameters to be estimated are for example effective processor speed, memory consumption and network throughput of the hardware platform on which software is deployed. In this paper we show how Multi-Paradigm Modelling (MPM) allows for the synthesis of a model calibration infrastructure. This includes the synthesis, from a model, of a simulator for the "environment" in which a system-to-be built will operate. To demonstrate the feasibility of our approach, we calibrate the model of an automotive power window controller running on the AUTOSAR-platform.