Multiuser multirate DS-CDMA detection techniques in multipath fading channels

Code division multiple access (CDMA) is a promising technique for wireless communication where multiple users share the same channel, but is subject to intersymbol interference (ISI), multiple access interference (MAI) and near-far effects under multipath channels. This thesis investigates various multirate multiuser detection techniques (MUD) to mitigate these interferences in multirate Direct Sequence (DS) CDMA communication systems over multipath Rayleigh fading channels.; A new decorrelator based successive interference cancellation (DBSIC) multiuser Rake receiver is proposed for asynchronous Variable Processing Gain (VPG) multirate DS-CDMA systems. By including a decorrelator on top of matched filtering (MF) at each stage of the conventional SIC to determine the user's signal to be used at the next cancellation stage, DBSIC improves the overall system performance at the expense of feasible additional complexity.; Bit-error-rate (BER) performance of DBSIC is presented using an analytical approach and simulations. DBSIC improvement over other commonly used suboptimum MUD such as the decorrelating, minimum mean-square-error (MMSE), SIC, parallel interference cancellation (PIC) and decorrelating decision-feedback (decorrelating DF) detectors is demonstrated under both perfect and imperfect channel estimation.; It is shown that DBSIC outperforms all other considered MUD in various multi-rate scenarios even when some physical users have increased data rates (i.e. heavier system's load in terms of the number of virtual users). It is observed that, while all the receivers suffer performance degradation under imperfect channel estimation, DBSIC still outperforms the other schemes. DBSIC effectiveness to combat MAI is also evaluated in terms of its asymptotic multiuser efficiency (AME). AME for DBSIC is derived and compared with that of conventional single-user MF, optimum, decorrelating and SIC receivers over AWGN and Rayleigh channels.; To combat effectively the effects of multipath propagation and MAI, most MUD require some channel state information (CSI). Their performances depend mainly on their ability to obtain an accurate estimation of the CSI they require. An adaptive MAI-resistant joint channel estimation and MUD method combining decision-directed and data-aided approaches is developed for asynchronous multirate CDMA systems. The effectiveness of this joint scheme is supported by results in terms of channel estimates MSE and BER performance.