The Algorithm Research And System Simulation Of Blind Signal Detection And Processing

Blind signal processing (BSP) is the new field of research which has been developing since the latest decade of the 20th century. It means separating or estimating the waveforms of source signals from the outputs of the system in the case of unknown of transmission channel characteristics and inputs information or only a small amount of priori knowledge. BSP contains blind source separation (BSS), blind signal extraction (BSE), blind deconvolution, blind equalization, blind identification and so on.Blind signal processing has become one of the hot topics of signal processing at present and has shown an attractive application prospect in fields of seismic exploration, mobile communication, voice processing and biomedicine engineering and so on. Among them, sound source blind signal detection and processing, especially voice processing, has become one of the important application areas of blind signal processing.This paper focuses on the main algorithm system of BSS, independent component analysis (ICA), researches and realizes a variety of important ICA algorithm, including switching algorithm of blind source separation based on kurtosis, FastICA algorithm, the algorithm based on global optimal property, the blind source separation algorithm based on maximum signal noise ratio and second order blind identification algorithm and so on. At the same time this paper improved the blind source separation algorithm based on maximum signal noise ratio which improves the performance of the algorithm and solves the problem of losing the effect under some conditions of the original algorithm.And blind signal processing is applied to sound sources in this paper. A sound source blind signal detection and processing system was developed with integrating the above algorithm into it. The whole process of mixing and separating sound source blind signal and the relatively complete researches of sound source blind signal detection and processing can be achieved through this system which provides a relatively comprehensive research platform for this field. A series of experiments were carried out using this system and many important conclusions were learned through comparing and analyzing experimental results, which had some guiding significance for actual sound source blind signal detection and processing.