Abstract:

As the parameters estimation of multiple direct sequence spread spectrum (DSSS) signals in noise condition is difficult for traditional methods, the denoising source separation (DSS) method is introduced to achieve the goal of mixedsignal separation and noise reduction. First parameters estimation of the DSSS signal based on cyclic spectrum techniques is reviewed under the conditions of low SNR. And the blind source separation algorithm is explained to be feasible in mixed DSSS signal parameters estimation. Then the basic model of blind source separation and an effective algorithm called biasremoval fast independent component analysis (FASTICA) are given. An algorithm framework for blind sources separation—DSS is derived, and that classical ICA algorithms can be unified into this framework is proven. The simulation results show the effectiveness and practicability of the algorithm.