Analysis of Norms in Adaptive Algorithm on Application of System Identification

System identification is an important area in signal processing research. It aims to retrieve the system’s unknown specifications from its output only. This technology has a wide variety of applications in engineering and control, industries, as well as medical fields. Typically, the identification of models expressed as mathematical equations. Linear, Non-Linear, Non parametric and Hybrid models are few deciding factors on which different techniques for System Identification relies on. In this paper, we discuss in detail the LMS algorithm and NLMS algorithm. In particular, various types of norms are included in LMS algorithm and the NLMS algorithm is modified according to the norms. Considering different norms in LMS algorithm we have analyzed the application of System identification. Also, it has been verified for both linear and non-linear models. Finally, for non-linear system identification based on Wilcoxon norm has been proposed. The results as well as the comparison show that the Wilcoxon norm is one of the better norms than others and is applied for System identification. The results show its efficacy.