Accurate Spectral Analysis of Repetitive Signals Composed of Components Nonsynchronous With the Sampling Clock

A method for the accurate spectral analysis of signals containing components that do not require synchronism either with each other or with the sampling clock is presented. The method employs an algorithm based on the quadrature amplitude demodulation which is applied in a number of stages to achieve an accurate decomposition of a given time domain signal into frequency components. Using the simulated signal examples, the suitability of the method for the analysis of noisy quantized nonrepetitive signals is considered. When compared to the discrete Fourier transform, the benefits of using the method for the accurate detection of nonsynchronous tones in a signal are demonstrated