Robust recursive spectral estimation based on an AR model excited by a t-distribution process by using QR decomposition algorithm

In this paper a new robust recursive, QR decomposition based, spectral estimation which is based on an AR model is proposed. The parallelism of the QR decomposition approach is used to facilitate the possibility for implementing the algorithm on an array processor architecture. The optimal coefficient of the AR model is selected by assuming that the excitation signal is a t-distribution with a degrees of freedom. When /spl alpha/=/spl infin/, we get the conventional QR decomposition RLS method. Simulation results show that, when the excitation signal is spiky, the obtained estimates using the proposed method with small /spl alpha/ are more efficient, the standard deviation (SD) of the estimation results are smaller, and more accurate than that with large /spl alpha/ and with Huber's estimator.