The adaptive correction problem of preliminary processing of the signals observed on noise background in the process of spectral analysis of the signals by Root-MUSIC method when using the surrogate data technology, obtained by randomization of observation component phase. The simulation results of the adaptive correction of observation processing are presented that show the efficiency of application this technology for the increasing of estimation accuracy of components of signal power spectrum in the case of low signal-to-noise ratios (SNR). The estimation accuracy obtained with using the adaptive correction as opposed to nonadaptive correction in the case of high SNR is compatible with accuracy of initial Root-MUSIC method.
surrogate data, eigenstructure methods, adaptive correction
"Adaptyvnaia korrektsyia predvarytelnoi obrabotky syhnalov s yspolzovanyem tekhnolohyy surrohatnыkh dannыkh v zadachakh spektralnoho analyza" ,
Information Processing Systems,