Joint iterative detection for random coded code-division multiple access

Joint iterative decoding of multiple forward error control (FEC) encoded data streams is studied for linear multiple access channels, such as code-division multiple-access (CDMA). Such systems can be viewed as serially concatenated communication links, and iterative turbo decoding principles can be successfully applied. Iterative decoders are made up of two basic components, a front-end processor that resolves the multiple-access interference and a bank of parallel error control decoders.; To improve power efficiency at low signal-to-noise ratios, powerful FEC codes, which are themselves serially concatenated, are used. Therefore the overall transmission system is a cascade of two error control codes with the CDMA channel, and soft-decision decoding is performed at each stage.; We conduct an empirical analysis of these component systems, using an error variance measure of the soft symbols generated by the soft-input soft-output (SISO) component decoders. All these soft symbols are derived from the extrinsic information according to the turbo decoding rule. It is shown how a variance measure leads to a variance transfer (VTR) analysis between component systems, which allows very accurate predictions about the convergence capability of the joint iterative decoder proposed.; This technique is used to study the effects of linear multiuser detectors such as simple interference cancellation, decorrelator and minimum mean-square error (MMSE) filter in the joint iterative receivers. Closed-form VTR functions are derived for these linear multiuser detectors as the number of users tends to infinity.; Guidelines for choosing error control codes that are suitable with respect to different channel noise level and system loads are proposed based on VTR methodology to optimize the spectral efficiency of the transmission link. Design examples are presented and put in contrast with the theoretical capacity limits of the random CDMA channel.