A DSP-based DS-CDMA multiuser receiver employing partial parallelinterference cancellation

摘要

The implementation of advanced DS-CDMA receivers based onnmultiuser detection principles is becoming a reality thanks to thencombination of an improved understanding of the theoretical basis ofnmultiuser detection and advances in digital, mixed-signal, and RFntechnologies. Due to their lower complexity, subtractive interferencencancellation approaches are attractive for the practical implementationnof multiuser detection. In a parallel interference cancellationnreceiver, it is practical to use the soft outputs of a matched filternbank for amplitude estimation. A bias arises in the decision statistics,nhowever, due to imperfect estimation and interference cancellation. Innthis paper, the source of the bias is explicitly recognized, and anpartial interference cancellation scheme that mitigates the negativeneffects of biased estimation and significantly improves systemnperformance is proposed. A practical real-time algorithm thatnsignificantly reduces the implementation complexity of this schemenwithout sacrificing performance is then derived. To facilitate ansoftware radio implementation, the signal processing complexity of thenapproach is characterized. The real-time processing algorithm is testednvia implementation in software on a floating-point general-purpose DSP.nThe prototype includes a flexible software-based architecture whichnperforms IF sampling and uses digital downconversion prior to basebandnprocessing. The hardware test setup is described, and the results arenpresented and compared with simulation and analytical results. Thenexperimental results confirm the simulation and analytical results whichnshow large performance gains over the conventional matched filter