Cascade Structure for Finite Impulse Response Filters and Linear Prediction

Abstract

This paper presents a complete analysis of the cascade structure for adaptive transversal filters based on adaptive algorithms. The standard structure of the cascade transversal FIR filter is obtained by replacing the whole structure by small ones with the same impulse response but having a less number of taps than the original structure. Computer simulation result shows the validity, reliability and the limitations that the model could have in its capacity of prediction. The optimal values of the model are compared with those obtained by the standard least mean square and recursive least square adaptive algorithms in order to verify the convergence of the weights and determine how fast this structure achieves those weights. For this case the speed of the algorithm is determined by the number of iterations that the filter requires to reach the minimum square value of its learning curve.