QFM Signals Parameters Estimation Based on Double Scale Two Dimensional Frequency Distribution

Abstract

This paper proposes a novel estimation method, named double scale two dimensional frequency distribution (DSTFD), to estimate the parameters of quadratic frequency modulated (QFM) signals. In the DSTFD, by using a novel parametric instantaneous self-correlation function and the idea of the keystone transform, the QFM signals are transformed into two-dimensional frequency domain, and the QFM signals are detected by searching peaks. To reduce computational cost, a double scale (DS) estimation strategy, which consists of coarse estimation and fine estimation, is proposed. The DS estimation strategy can be implemented by using the Chirp Z-transform and an improved transform named local scaled Fourier transform (LSFT). The LSFT only consists of complex multiplications, fast Fourier transform (FFT), and inverse FFT operations. The implementation, anti-noise performance, and computational cost are analyzed for the proposed method. Through simulations and analyses, the results demonstrate that the DSTFD outperforms other compared algorithms. INDEX TERMS Double scale two dimensional frequency distribution (DSTFD), quadratic frequency modulated (QFM), parameter estimation, parametric instantaneous self-correlation function (PISCF), double scale estimation strategy.

Lower Limb Analysis Based on Surface Electromyography (sEMG) Using Different Time-frequency Representation Techniques

“…Moreover, performing the classification via 2D images format has reached higher than using the 1D time-domain signal. Thus, mapping the 1D sEMG signal in time series into 2D space can help explore all discriminatory features in both frequency and time domains [11], [16], [22], [25], [27]. The knee muscle time-domain segments were mapped into 2D time-frequency space using three timefrequency representation techniques.…”

“…1) The scalogram is generated by continuous wavelet transform (CWT). CWT preserves time shifts and time scales in terms of time-frequency resolution [11]. Wavelet transform can analyze the sEMG signals in the time and frequency domain and defined by the mother function given by (1) [26]:…”

“…2. The persistence spectrum is computed based on the time percentage that a given frequency persists within the signal [11]. It is a histogram-based spectrum represented by power frequency [12].…”

LFM Signal Analysis Based on Improved Lv Distribution

“…It is known that the similar coupling can be removed by using the keystone transform [30]. Inspired by the idea of keystone transform, we propose the ISFT-IFT operator in [36] which is expressed by…”

“…In the ImLVD, the two-scale estimation strategy is used to reduce the computational cost. To implement the strategy, the Chirp-Z transform (CZT) [35] and the improved scaled Fourier transform (ISFT) [36] are used. In the ISFT, by introducing a scaling factor, a new time variable is constructed to remove the coupling.…”