Iterative PDF estimation and turbo-decoding scheme for DS-CDMA systems with non-Gaussian global noise

Abstract

In some multiuser direct-sequence code-division multiple-access (DS-CDMA) systems, the commonly used Gaussian approximation (GA) is inaccurate to describe the probability density function (PDF) of the global noise, which is the sum of the multiple-access interference (MAI) and channel noise. This inaccuracy causes the performance degradation of turbo codes as compared to decoders based on the true PDF. Therefore, an iterative PDF estimation and turbo-decoding scheme is proposed, where the kernel-based method is used to estimate the PDF from the global-noise samples refined from the received signals, and the information is iteratively exchanged between the PDF estimator and the turbo decoder. This scheme achieves performance comparable to decoders based on true PDF after several iterations.