Fast adaptation of equalizer and desired impulse response for digital data receivers

The paper considers the design of the equalizer and the desired overall system impulse response (d. i. r. ) for various types of detection algorithms used to recover digital data which has suffered intersymbol interference. The joint optimization of the equalizer and desired response is formulated using a mean-square error cost. A constraint on the desired response is introduced which enables simple adaptive algorithms to be derived for determining the optimum equalizer and d. i. r. One of the algorithms considered has been found to converge rapidly and is thus ideally suited for application where fast start-up is required. Simulation results of the probability of error performance of various detection algorithms using the equalizer and d. i. r. designed according to the approach developed in the paper, are presented. These results show that the performance obtained is comparable to that achieved with systems in which the equalizer and d. i. r. are designed according to other techniques. The disadvantage of these other techniques is that they usually lead to algorithms which are difficult to make adaptive and are slow to converge.