Proposal of a novel signal separation principle based on DFT with extended frame Fourier analysis

An analytical principle for a time limited signal is made by introducing the extended frame concept into DFT (discrete Fourier transform). Waveform analysis more accurate than DFT is achieved by calculating correlation between the input frame signal with an extended frame and the element frequency components corresponding to the extended frame length. It is shown that this method contributes to enhancing functions such as precise filtering, signal separation, or noise suppression. The improvement of signal to noise ratio for FSK/PSK data transmission is estimated, compared to the performance with conventional matched filter detection