Blind channel equalization and identification over frequency selective slow fading channels

A combined recursive zero-forcing and least squares (RZF-LS) blind algorithm is presented. The method employs the statistical and the structural characteristics of the signal matrix to conquer the sensitivity of channel order estimation errors inherent in some existing blind algorithms. An accurate channel order detection algorithm in low SNR is promoted based on the analysis of the statistical characteristics of the input signal. Numerical simulations show that the proposed RZF-LS blind algorithm leads to a more accurate and robust estimate in channel estimation and bit error rate (BER) performance.