System identification using LMS, NLMS and RLS

In this paper system identification has been done using adaptive filters. System identification is the process of identifying an unknown system form input output signal. It can be defined as the interface between real world of application and mathematical world of control theory and model abstraction. Three types of adaptive filters are used to identify the unknown system Least Mean Square (LMS), Normalized Least Mean Square (NLMS) and Recursive Least Square (RLS) algorithms. LMS has less computational complexity than NLMS and RLS while NLMS is the normalized form of LMS adaptive filter. RLS is complex algorithm but it works more efficiently. All these algorithms works on the basis of Least Mean Square Error (LMSE) and filter's weights are recursively updated as to bring output signal equal to the desired signal. These algorithms are applied to the unknown system and the simulation results are compared.