Blind equalization and multiuser detection for high rate CDMA communications

The problem of blind demodulation of the users' information symbols in a high-rate CDMA network with the presence of both multiple-access interference (MAI) and intersymbol interference (ISI) is considered. The dispersive CDMA channel is first cast onto a multiple-input multiple-output (MIMO) model. By applying the theories of blind channel identification and symbol detection in MIMO channels, under certain conditions, the multiuser information symbols can be recovered without any prior knowledge of the channel or the users' signature waveforms (including the signature of the user of interest), although the algorithmic complexity of such an approach is prohibitively high. However, in practice, the signature waveform of the user of interest is always available at the receiver. It is then shown that by incorporating this knowledge, the impulse response of each user's dispersive channel can be identified using a subspace method. It is further shown that based on the identified channel response, both the decorrelating multiuser detector and the linear MMSE multiuser detector can be constructed in closed-form, at almost no extra computational cost. Data detection can then be furnished by applying these linear multiuser detectors (obtained blindly) to the received signal.