Improved Multiuser Detection in Asynchronous Flat-Fading Non-Gaussian Channels

In this paper, a new M-estimator based multiuser detection in asynchronous flat-fading non-Gaussian CDMA channels is considered. A new closed-form expression is derived for the characteristic function of the multiple-access interference signals. Simulation results are provided to prove the effectiveness of the derived bit-error probabilities obtained with this expression in asynchronous flat-fading non-Gaussian CDMA channels.

💡 Research Summary

The paper addresses the long‑standing challenge of reliable multi‑user detection (MUD) in code‑division multiple‑access (CDMA) systems when the channel is both asynchronous and subject to flat fading, and when the additive noise deviates from the Gaussian assumption. Traditional linear detectors such as the minimum‑mean‑square‑error (MMSE) filter or the decorrelator, as well as conventional maximum‑likelihood (ML) approaches, are known to perform poorly under impulsive or heavy‑tailed interference because they rely heavily on second‑order statistics and Gaussian noise models. To overcome these limitations, the authors propose a robust M‑estimator‑based detector that replaces the conventional quadratic loss with a more general, bounded loss function ρ(e) and its derivative ψ(e). By choosing ψ functions derived from Huber’s or Tukey’s bi‑weight criteria, the detector automatically down‑weights large residuals that are likely caused by impulsive noise components, thereby achieving strong robustness without the need for explicit noise modeling.

The system model assumes K users transmitting binary symbols (±1) that are spread by user‑specific spreading sequences and experience independent, flat Rayleigh fading coefficients hk(t) together with random propagation delays τk. The received baseband signal is a superposition of all users’ faded, delayed waveforms plus a composite noise term n(t). The noise is modeled as a Gaussian‑Mixture (GM) process: a dominant low‑power Gaussian component combined with a sparse high‑power impulsive component, a realistic representation of many wireless environments where man‑made interference or hardware non‑idealities generate occasional spikes.

A central theoretical contribution of the work is the derivation of a closed‑form expression for the characteristic function (CF) of the multiple‑access interference (MAI) under the above conditions. Unlike prior work that approximates MAI as Gaussian by invoking the central limit theorem, the authors retain the exact statistical dependence on the asynchronous delays, fading amplitudes, and spreading codes. By applying Fourier transform techniques and exploiting the independence of user symbols, they obtain a product‑form CF:

\