FSK signals detection by digital filtration based on capon approach

In communication and RF-monitoring first step of received signal processing is to determine start moment of signal which contains useful information. This processing can be made on basis of digital linear band filtration. In this paper we describe algorithm of digital harmonic fill filtration based on spectral Capon approach. Received vector of filter coefficients is the solution of minimum dispersion problem for signals with a singular autocorrelation matrix. Algorithm energy efficiency has been investigated in the problem of FSK signals detection in presence of additive noise and Doppler effect.