Robust Multiuser Detection in Synchronous Chaotic Modulation Systems

Abstract

In reality, many wireless channels exhibit marked departure from Gaussian noise assumption, causing significant performance loss when linear multiuser detectors, which are optimum in the case of Gaussian noise, are used for detection of code division multiple access signals. In this article, we propose a new influence function-based M-estimator and its application to the problem of robust multiuser detection of code division multiple access signals in chaotic modulation systems, in which user data are modulated by an analog chaotic series rather than periodic binary spreading sequences, against non-Gaussian background noise. The solution to this robust detection problem is not in closed form and requires an iterative procedure, which is a limitation in its practical use. However, it causes a little attendant increase in computational complexity in practical applications. Robustness analysis and simulation results show that the proposed detector offers significant performance gains over least-squares and Huber M-estimator-based detectors in chaotic modulation systems in non-Gaussian noise.