Multiuser detection with partial information for asynchronous CDMA-based radio networks

We propose and evaluate a form of multiuser detector for base station reception in CDMA wireless. The setting we have in mind is one in which the interference at any base station has components whose parameters-power delay, signature sequence-are known to the receiver as well as components, representing out-of-cell transmissions, for example, whose parameters are unknown. The signals to be jointly decoded are thus to be extracted from a lager aggregate, plus noise, on the basis of partial parameter information. The setup provides a framework in which to study the impact of parameter information and detection group size on receiver performance. The proposed receiver architecture, amenable to adaptive as well as non-adaptive implementation, features a bank of linear equalizers at the input and a maximum-likelihood detector at the output; performance is described in terms of the mean-square error bit error rate and asymptotic efficiency. The computational complexity per bit, given the size of the detection group, is independent of the number of interferers.