Fast ICA and JADE Algorithms for DS-CDMA

Abstract

This paper introduces application of Independent Component Analysis (ICA) in communication field for Direct Sequence Code Division Multiple Access (DS-CDMA) in symbol estimation, where ICA is statistical method for source separation based on independence between sources, and the paper developed a way to solve the ambiguities of ICA. We used ICA as a detector and it is simulated using MATLAB in downlink DS-CDMA. Fast ICA and JADE algorithms which are ICA algorithms are introduced for symbol detection for different number of users and different number of symbols. A comparison between the two algorithms is also made, the results are compared with results of conventional detection The results showed that ICA detection can solve symbol detection. The ICA detection after solving ambiguities using the proposed way gives better results than conventional detection