Performance of adaptive linear interference suppression for DS-CDMA in the presence of flat Rayleigh fading

The performance of adaptive linear interference suppression for direct-sequence code-division multiple-access is evaluated in the presence of flat Rayleigh fading. Interference suppression is achieved with an adaptive digital filter which attempts to minimize the mean squared estimation error. A model for the reverse link of a cellular system is assumed which accounts for both intra- and other-cell interference, and stochastic arrivals and departures of transmitted packets. All received signals experience both long-term shadowing and short-term Rayleigh fading. Variations on decision-directed and blind least squares adaptive algorithms are presented which are compatible with differential coding and detection. Computer simulation results are presented which compare the uncoded performance of the adaptive scheme with the conventional matched filter. These results indicate that the adaptive approach is insensitive to variations in received power (averaged over fades), and offers a significant gain in capacity (measured in Erlangs per cell) at modest to high error rates (5-15%).