Degradation of the tracking performance of adaptive filtering algorithms with data correlation (Q2732949)

The paper provides a quantitative evaluation of the degradation of the performance of various adaptive filtering algorithms with data correlation. The algorithms examined are the least mean square, the recursive least squares, the sign, the signed regressor and the sign-sign algorithms. The analysis is done in the context of the identification of a randomly time-varying plant and is based on the degradation of both the steady-state excess mean-square error and the steady-state mean-square weight misalignment.