Unbiased FIR system identification in the presence of input and output interference

In the presence of input interference, the Wiener solution for impulse response estimation is biased. In this paper, it is proved that bias removal can be achieved by proper scaling of the optimal filter coefficients and a modified least mean squares algorithm is then developed for accurate system identification in noise. Simulation results are included to compare the impulse response estimation performances of the proposed method and two total least squares based adaptive algorithms under different interference conditions.