An effective design of multistage parallel interference cancellation (PIC) receiver using blind adaptive (BA) despreader and pre-respreader interference estimator for uplink CDMA is proposed and analysed. A novel algorithm is designed, which exploits constant modulus (CM) property of the users' transmitted signals and inherent channel condition to perform adaptive despreading based on minimum error variance criteria. This is carried out by BA weighting of each chip signal for accurate tracking of the desired user's signal power and hence for more improved data detection at the output of each stage of PIC. Furthermore, the despreader weights are used within the adaptive pre-respreader interference estimation and cancellation to obtain online scaling factors during every symbol period, without any knowledge of users' channels or the use of training sequences. It is found that this way of estimation is optimal in minimum mean, squared error sense, and hence, significant reduction in interference and noise variance is observed in detection and estimation of the desired users' signals compared with conventional PIC. Bit error probability of the proposed PIC is obtained using Gaussian Approximation method. Extensive simulation results are shown, which demonstrate impressive performance advantage in fading environments, high system loading, and severe near far conditions.
- Blind channel estimation
- Parallel interference cancellation