Skip to Main Content
A blind MMSE equalizer based on linear prediction is presented in this paper. It can directly equalize the channel from the fractionally sampled observations without performing channel identification. In addition, it can be implemented efficiently using the RLS algorithm. Simulation results show that the blind MMSE equalizer has smaller mean-square error of symbol estimation than the prediction-based zero-forcing equalizer.
Date of Conference: 31 May-2 June 2004