Performance Enhancement of Downlink Multiuser DS-CDMA Detectors Using Processing by Independent Component Analysis

In this paper, the attention is focused on the generation of the received digital signal and its linear detection over a synchronous multiuser direct sequence code division multiple access (DS-CDMA) downlink system considering various spreading codes, additive white Gaussian noise (AWGN) and introducing varying signal-to-noise ratios (SNRs) and users' power distributions. The aim of the work is to test the efficiency of implementing an independent component analysis (ICA) based estimator of the transmitted symbols preceded by principal component analysis (PCA) pre-processing of the received data to reduce dimension and decrease the noise effects. The noisy ICA algorithm is used either as: (1) a pre-processor prior to conventional detection: The idea here is to estimate the mixing matrix that contains the basic vectors and fading terms using the separating matrix and PCA subspace or as (2) a post-processor attached to a linear DS-CDMA receiver. Numerical simulations indicate that the first schema could constitute a powerful tool for channel and symbols estimations and has the advantage that it doesn't necessitate any prior knowledge of the users' codes or paths delays and strengths except providing a short pilot sequence (presenting 1% to 4% of the transmitted symbols of the user of interest) that is used to give a good enough initial guess of the demixing matrix columns and so force the ICA iteration to be in the wanted user subspace. The second schema shows performance comparable to or superior than the exact blind detectors at relatively high SNR values.