The evaluation of saturation level for SMSE cost function in identification of elementary bilinear time-series model

It is possible to obtain an unbiased coefficient estimates of a stable elementary bilinear time-series model by using Saturated Mean Square Error as a cost function of identification algorithm. However, a proper evaluation of Saturation Level is required to ensure correct placement of the global minimum of the cost function therefore, the variance of the original innovation signal is required. Due to lack of this information before the identification run, an approximation procedure is necessary which is proposed in this paper.