Research On Key Technologies In Blind Signal Processing

With the ever-changing communication technology,the fields involved are also becoming wider and wider.Blind signal processing technology is an emerging signal processing technology.Blind detection is to detect whether there is an information signal from the observed signal,Signal pattern recognition is a modulation mode that obtains the original signal from the obtained communication signal,Blind source separation and blind signal extraction are effective methods to recover the source signal from the observed signal.These blind signal processing techniques have become a research hotspot in the field of signal processing.Blind signal processing technology has a wide range of applications in communication systems.For example,in the field of military communications countermeasures,electronic interference,etc.,radio monitoring over civilians,etc.This paper first studies a blind detection method based on classical energy detection algorithms.The detection performance of classical energy algorithms is affected by the noise variance of the channel,The noise variance of the channel is obtained through estimation,If the estimation error is large,it will directly affect the detection probability,thus making the detection inaccurate.The blind signal detection proposed in this paper is not affected by the channel noise,which makes the detection accuracy much better than the classical energy algorithm.Simulation experiments show that this blind spectrum detection algorithm has a better detection effect than classical energy detection methods and does not require known noise variance information.Afterwards,theoretical studies were conducted on the differences of several common digital modulation signals in the cyclic spectrum curves,The characteristics of the cyclic spectrum function of digital modulated signals such as MASK,2FSK,MFSK,BPSK,and QPSK are analyzed,Through simulation experiments,it can be shown that the common digital modulation signal can be judged by cyclic spectrum.Finally,the basic principle of blind source separation and the classic FastICA algorithm are analyzed.The FastICA algorithm completes the whitening process of the data,and iteratively calculates the separation matrix to complete the blind separation of the mixed signals.Simulation experiments show that under the noise-free condition,FastICA algorithm can separate different independent source signals.However,under noisy conditions,the separation effect will increase with the increase of signal-to-noise ratio.