Application of Correlation Analysis in Weak Signal Detection

摘要

In paper the problem of detection a low-frequency weak signal from the mixture with strong random noise is considered. It is shown that usage of correlation analysis makes it possible to detect a harmonic signal from an additive mixture of a useful signal and random noise. But with a single applying of correlation, this information can be lost against a background of strong noise. To solve this problem, it is recommended to repeatedly determine the correlation function. In this case, the correlation function of the mixture becomes equal to the correlation function of the useful signal. As a result, the signal-to-noise ratio increases many times. Also, based on the obtained autocorrelation function, it is possible to estimate the period and amplitude of the harmonic weak signal.