Remove multifrequency mixture noise in processing of Wireless Communication wave/radar reflect signal by method of ICA

The ICA technique is used to remove multifrequency mixture noise in Wireless Communication wave/radar reflects signal processing. The characteristics of Wireless Communication wave/radar reflect signals are analyzed, build up models of expectation signal and consociation mixture noise and a constituted method of the independence prognosticate machine is discussed. Use the improved algorithmic of FastICA to simulate it. From simulation results, a well effect is acquired by method of the mixture signals separate above-mentioned. Via this, it can gain source signals almost pure furthermore wider the range ratio scope of each signal channel and type of signal. The method in the text possesses very good anti-jamming ability and it is can be expand other related signal processing, so the method is widely accepted.