Blind matched filter receiver

Baykal B.

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS I-REGULAR PAPERS, vol.51, no.2, pp.379-390, 2004 (SCI-Expanded) identifier identifier


Matched filter receiver (MFR) of Forney in the presence of intersymbol interference (ISI) and additive white Gaussian noise is investigated from a blind perspective, which includes blind maximum likelihood sequence detection, blind decision feedback equalization and linear blind equalization. Various realizations of the blind MFR are discussed. In a blind MFR, the channel can be easily estimated using a simple algorithm such as the constant-modulus algorithm (CMA). In this sense, the CMA can be used to estimate the channel directly without complicated matrix decomposition techniques. Zero forcing and minimum mean-square estimation based designs of the blind MFR are explored. Channel estimation using the CMA deteriorates in the presence of noise. A new method to tackle this problem is developed and shown to have satisfactory behavior in low signal-to-noise ratios. Tracking performance of the blind MFR is found to be acceptable in representative Rayleigh fading channels of mobile communication.