2026-04-17T17:41:01.870Z http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai

ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai:2:186 2013-10-21T19:41:39Z 2

Robust adaptive filtering with variable forgetting factor 2004 http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai/record/2/186 Z. Banjac B. Kovačević M. Veinović M. Milosavljević In this paper, a new robust adaptive FIR filter algorithm with variable forgetting factor for time-varying parameters estimation in the presence of impulsive noise is discussed. The proposed algorithm is based on the robust recursive least square and parallel adaptation recursive least square algorithms combination. A scalar parameter controls the switching process between these two algorithms, and relates to the decision that the instantaneous desired response of the adaptive filter does not contain significant impulsive noise. This parameter calculation is based on the median and mean operation at the error signal samples from the data window of a proper size. The feasibility of the approach is demonstrated in a system identification simulation using a FIR filter application. article 3 2 223 228 1109-2734 WSEAS Transactions on Circuits and Systems