Machine Learning and Intelligent Communications. Second International Conference, MLICOM 2017, Weihai, China, August 5-6, 2017, Proceedings, Part II

Research Article

Parametric Sparse Recovery and SFMFT Based M-D Parameter Estimation with the Translational Component

Download
103 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-73447-7_16,
        author={Qi-fang He and Han-yang Xu and Qun Zhang and Yi-jun Chen},
        title={Parametric Sparse Recovery and SFMFT Based M-D Parameter Estimation with the Translational Component},
        proceedings={Machine Learning and Intelligent Communications. Second International Conference, MLICOM 2017, Weihai, China, August 5-6, 2017, Proceedings, Part II},
        proceedings_a={MLICOM},
        year={2018},
        month={2},
        keywords={Compressive Sensing (CS) micro-Doppler effect (m-D effect) Parametric sparse representation Sinusoidal Frequency Modulated Fourier Transform (SFMFT) 
            -resolution Fourier-Bessel (-FB) series Parameter estimation},
        doi={10.1007/978-3-319-73447-7_16}
    }
    
  • Qi-fang He
    Han-yang Xu
    Qun Zhang
    Yi-jun Chen
    Year: 2018
    Parametric Sparse Recovery and SFMFT Based M-D Parameter Estimation with the Translational Component
    MLICOM
    Springer
    DOI: 10.1007/978-3-319-73447-7_16
Qi-fang He1,*, Han-yang Xu2, Qun Zhang1, Yi-jun Chen1
  • 1: Air Force Engineering University
  • 2: Xidian University
*Contact email: qifanghe@163.com

Abstract

The micro-Doppler effect (m-D effect) provides unique signatures for target discrimination and recognition. In this paper, we consider a solution to the m-D parameter estimation. This method mainly consists of two procedures, with the first being the radar returns decomposition to extract the m-D components in Bessel domain. Then the parameter estimation issue is transformed as a parametric sparse recovery solution. A parametric sparse dictionary, which depends on m-D frequencies, is constructed according to the inherent property of the m-D returns. Considering that the m-D frequency is unknown, the discretizing m-D frequency range for the parametric dictionary matrix is calculated by the sinusoidal frequency modulated Fourier transform (SFMFT). In this manner, the finer m-D frequency, initial phases, maximum Doppler amplitudes and scattering coefficients are obtained by solving the sparse solution of the m-D returns. The simulation results verify the effectiveness.