Frame-oriented channel estimation for UTRA/FDD with LSE polynomial fitting

Channel estimation is a crucial issue in CDMA systems and especially for UTRA/FDD. Many techniques have been proposed to perform enhanced channel estimation, in particular WMSA (weighted multi-slot averaging). This technique yields very satisfactory results, but it is not able to adapt to the speed of channel variations. We propose a frame-by-frame estimation technique. It determines a polynomial function fitting the noisy channel estimates in a least squares sense. Moreover it is able to adaptively choose the degree of the polynomial depending on the speed of channel variations. The obtained results show an SNR improvement up to 1 dB with respect to WMSA.