Performance analysis of the self-correcting adaptive filter

An adaptive noise canceller is a well-known method for removing noise from noise-corrupted speech. The problem arises in many situations such as airplane cockpits and automobiles. The purpose of the self-correcting adaptive filter (SCAF) method is to improve the accuracy of filters with small order compared to other methods. This paper extends the results of the SCAF by investigating the limit to which the SCAF algorithm approaches when large and abrupt changes of the desired signal as well as the noise signal are present. Simulation results are presented to compare the performance of the SCAF algorithm with the fixed step size LMS algorithm.