Researches On Algorithms And Applications Of Blind Signal Separation

Blind signal separation (BSS) is particularly important to radio management. In radio monitoring, the signal that observed by observation stations are mixed signals by many signals of the same frequency. We will use the theory of BSS to estimate the source signals.This paper researches on BSS theory thoroughly and systematically. These are the current situation,development and the theory of BSS. The description of theory includes the processing method and algorithms of BSS. the algorithms of blind signal separation were Simulated in this paper: algorithms based on maximun signal- to-noise ratio and the kurtosis with matlab to validate the two algorithms can separate blind signals.In this paper, an experiment was done to sample data. The experimental equipments are signal generators and a high-speed data acquisition card and so on. The algorithms simulated by matlab are applied to separate mixed-signals. The mixed-signals contains data are observed by the experiment and simulated by matlab. The separating effect is decided by similarity coefficient and the run time of the algorithms. The similarity coefficient is the level of similarity between the separated signal and the source signal.This paper did many simulations. Overall the tests, both of the two algorithms can separate the mix-signal successfully. But it appears that the separation effect by the approach based on maximun signal- to-noise ratio is better than the algorithms based on kurtosis. And the run time of the former is less than the latter.