SVD-Based Adaptive Multiuser Detection for Optimized Chaotic DS-CDMA Systems

In recent years, chaotic signals have created a new area in the designation of wideband communication systems. Most of the activity has focused on DS-CDMA systems, in which the conventional pseudo-noise sequences will be replaced by binary chaotic sequences. Unfortunately, despite the advantages of chaotic systems such as aperiodicity, low cost generation and noise-like spectrum, the performance of most of such designs is not still suitable for multiuser wireless channels. In this paper, we propose a novel method based on singular value decomposition for adaptive multiuser detection in chaos-based DS-CDMA systems. We also propose a new genetic algorithm-based method for the optimal generation of chaotic sequences in such systems. Simulation results show that our proposed nonlinear receiver with optimized chaotic sequences outperforms the conventional DS-CDMA systems with “maximal length” codes as well as non-optimized chaos-based DS-CDMA systems in all channel condition, particularly for under-loaded CDMA condition, which the number of active users is less than processing gain.