Multiple frequencies estimation of sinusoidal signals with sub-sampling

In some applications, it is hard to process high-frequency signals because of the constraint of A/D converter and latter signal processors’ low computation speed and high cost. A new method of multiple frequencies estimation of high-frequency sinusoidal signals based on sub-sampling is proposed. First use the power divider to make the high-frequency multiple sinusoidal signals into two pathway, then the two signals are sampled with different sampling rate respectively，using the Fourier transformation and improved Rife-Quinn algorithm to estimate the two separate frequencies. Then estimate the multiple frequencies unambiguously via spectral rapid-pairing utilizing the Chinese remainder theorem. This approach has the advantages of low cost, low sampling rate, high precision and easy implementation, even if in the very lower signal noise ratio (SNR). Step of this method is provided and simulations have verified the validity of the algorithm.