Enhancing the efficiency of spectral analysis of signals by the Root-MUSIC method using surrogate data

Signal processing correction in spectral analysis using the surrogate autocovariance observation functions obtained by the ATS-algorithm

“…However the saturation levels and their threshold SNR have smaller values as compared with the direct use of surrogate data obtained from observation. Figure 3 shows the similar characteristics for the case, where the correction of the Root-MUSIC method is performed by using the averaged surrogate (AS) realization [13] and averaged surrogate ACF. Figure 3 illustrates the correction of spectral analysis while using surrogate data obtained by the randomization of phases of observation Fourier-spectrum components (curve 3), randomization of phases of observation ACF Fourier-spectrum components (curve 4), and obtained by the ATS-algorithm from the observation ACF (curve 2).…”

“…This paper, as well as [15,16], employ just such variant of the algorithm. It should be noted that in the case of using the algorithm with randomization of the observation Fourier-spectrum phases, the dimension of the vector of surrogate data coincides with the dimension of the vector of observations [13].…”

“…One of the correction methods of incoherent spectral analysis of signals observed against the background of noise using the algorithm with randomization of the observation Fourier-spectrum phases is presented in [13].…”

“…The efficiency of processing correction of signals observed against the background of noise is shown in paper [13] for the case of signal incoherent spectral analysis by the Root-Music method using the surrogate data technology based on the algorithm with randomization of the observation Fourier-spectrum phases (random phase algorithm). The application of this technology is efficient both at small values of SNR and at large values of SNR, when the frequencies of signal components are multiple of the observation sampling frequency in frequency domain.…”