A FAMILY OF MULTIUSER DECISION-FEEDBACK DETECTORS FOR ASYNCHRONOUS CODE-DIVISION MULTIPLE-ACCESS CHANNELS

It is important to identify simple and reliable interference rejection methods for Code-Division Multiple-Access (CDMA) channels, since the conventional matched filter receiver has high error rates, and the optimal detector is too complex. We introduce decision-feedback and partial feedback detectors for asynchronous CDMA channels. Two-stage detectors with decision-feedback in the second stage are also studied. The derivation of the decision-feedback detector is based on spectral factorization which leads to a white-noise channel model. We also describe two implementations of the maximum-likelihood detector for this model. Comparisons among the proposed detectors, the conventional detector, and the linear decorrelating detector are undertaken for several asynchronous CDMA channels. In these examples, high bandwidth efficiency systems are explored, and user energies are varied from being similar to being very different. We find that decision-feedback detectors compare favorably with more complex two-stage methods and maintain good performance under diverse channel conditions.