Blind Estimation of PN Codes in Multi-user LSC-DSSS Signals

Abstract

Aiming at the problem of blind estimation of Pseudo-random (PN) sequences in multiuser long scrambling code direct sequence spread spectrum (LSC-DSSS) signals, in this paper, we proposed a novel PN sequences estimation method based on Fast-ICA algorithm and third-order statistics. The received signal is firstly divided twice into segments and then the segments are transformed into several array signal block matrixes based on the periodicities of scrambling codes and spreading codes. Every user's PN sequence is separated segment-by-segment using Fast-ICA algorithm, and those segments are assembled sequentially into fuzzy sequences. Finally, all the scrambling codes and spreading codes are estimated from the fuzzy sequences based on the triple correlation properties of m-sequences. The simulation results show the excellent performance of the proposed method.