DFT Spectrum-Sparsity-based Quasi-Periodic Signal Identification and Application

This study addresses identification of the quasi-periodic sequence based on the discrete Fourier transform spectrum-sparsity. A quasi-periodic signal displays similarities to a periodic signal but does not satisfy its strict definition. Then identification of the quasi-periodic signal is practically a requirement for many applications. The spectrum-sparse property in the Fourier domain is presented for describing the discrete-time periodic signals, and then it is extended and proposed for the quasi-periodic sequence identification. The proposed identification method is applied to real applications including periodicity detection of the sunspot activity, rainfall activity, and air temperature movement as well as the pile driver sound detection. The experimental results and analysis indicate potentiality of the proposed method that it yields satisfying results in the real applications.