Performance Comparisions of ICA Algorithms to DS-CDMA Detection

TL;DR

This paper evaluates and compares the performance of various ICA algorithms in improving multi-user detection in DS-CDMA systems, demonstrating their effectiveness over traditional methods.

Contribution

It provides a comparative analysis of major ICA algorithms for blind interference suppression in DS-CDMA, highlighting their impact on symbol estimation accuracy.

Findings

ICA algorithms improve interference mitigation in DS-CDMA

ICA-based detection outperforms traditional SUD in simulations

Combined SUD-ICA detection yields the best results

Abstract

Commercial cellular networks, like the systems based on DS-CDMA, face many types of interferences such as multi-user interference inside each sector in a cell to interoperate interference. Independent Component Analysis (ICA) has been used as an advanced preprocessing tool for blind suppression of interfering signals in DS-CDMA communication systems. The role of ICA is to provide an interference-mitigated signal to the conventional detection. This paper evaluates the performance of some major ICA algorithms like Cardoso's joint approximate diagonalization of eigen matrices (JADE), Hyvarinen's fixed point algorithm and Comon's algorithm to solve the symbol estimation problem of the multi users in a DSCDMA communication system. The main focus is on blind separation of convolved CDMA mixture and the improvement of the downlink symbol estimation. The results of numerical experiment are compared with those obtained by the Single User Detection (SUD) receiver, ICA detector and combined SUD-ICA detector.