Use of single bin Fourier transform algorithm for high speed tone detection and parallel processing

Abstract

Fourier transform is the classical tool for performing spectral analysis of a short-duration signal and fast Fourier transform (FFT) is the widely used algorithm for achieving the same with a high degree of computational efficiency. Recently a “single bin Fourier transform algorithm” has been proposed by the present authors, which is designed to compute the spectral amplitude and phase of a single frequency present in the time signal to be analyzed. The traditional detection scheme involves sampling the analog signal first using the Nyquist's sampling theorem, apply the FFT on the sampled signal and then identify the desired frequency from the FFT output. However, in case of tone detection, only the presence of a single frequency is to be tested. The computation of the whole FFT output dataset is superfluous in this case. In this paper, we have used the single bin Fourier transform algorithm for tone detection, which makes the detection scheme less complex and computationally more efficient than the traditional FFT algorithm. Moreover, the method of parallel processing can be applied to make the system more efficient in detecting the existence of more than one frequency component present in the incoming signal. The main objective of the proposed scheme is to improve the detection speed of a single frequency, or a narrow band of frequencies, present in a signal.