An improved Hammerstein-Wiener system identification with application to virtualized software system

This paper proposes a system identification procedure to approximate virtualized software system dynamics within the framework of a Hammerstein-Wiener model. The approach is an extension of the existing works where the linear dynamics are estimated in Frequency Sampling Filter (FSF) structure and the inverse static output nonlinearity are synthesized in B-Spline curve functions. Furthermore, the issue on parameter selection for B-spline model approximation is addressed by using a data clustering method. An experimental test-bed of virtualized software system is established to generate experimental data which are used to confirm the performance of the proposed approach. The identification results have shown that the model efficacy is increased with the proposed approach because the dimension of the nonlinear model is reduced significantly while maintaining the desired accuracy.