A subspace method for signature waveform estimation and detection in dispersive CDMA channels

We consider the problem of signature waveform estimation and detection in a dispersive CDMA channel in the presence of multiple access interference (MAI). A subspace training method for signature waveform estimation is proposed. By using this method the signature waveform can be estimated given only a small number of training symbols of the desired user. Simulation results show that the proposed method offers significant performance gain over the conventional matched filter receiver and it is near-far resistant.