The Experts below are selected from a list of 198 Experts worldwide ranked by ideXlab platform

Ljubiša Stanković - One of the best experts on this subject based on the ideXlab platform.

  • Micro-Doppler removal in radar imaging in the case of non-compensated rigid body acceleration
    2018 23rd International Scientific-Professional Conference on Information Technology (IT), 2018
    Co-Authors: Miloš Brajović, Ljubiša Stanković, Miloš Daković
    Abstract:

    The micro-Doppler (m-D) effect, commonly caused by fast moving reflectors, can significantly decrease the readability of rigid body in ISAR/SAR radar images. We revisit an L-statistics based micro-Doppler removal approach, producing excellent results in separation of the stationary rigid body from the m-D. In the case of non-compensated target acceleration, rigid body components become non-stationary, commonly with linear frequency modulation. The Local Polynomial Fourier Transform (LPFT) can be exploited for the estimation of the unknown chirp-rate needed for the acceleration compensation. To this aim, we present a simple iterative procedure based on the LPFT Concentration Measure. It is an alternative to the direct search approach for the estimation of the LPFT demodulation parameter, improving the estimation accuracy and reducing the numerical complexity of the approach. Numerical examples verify the presented theory.

  • Windowing methods for graph signal localization
    2017 22nd International Conference on Digital Signal Processing (DSP), 2017
    Co-Authors: Miloš Daković, Ljubiša Stanković, Ervin Sejdić
    Abstract:

    In this paper we considered windows used for local vertex spectrum analysis of graph signals. In addition to a review of the convolution based windowing method, two methods based on the vertex neighborhood are presented. They are based on the graph path lengths. In the first one the number of edges in a path determine window size, while the edge weights are taken into account in the second method. Signal localization is performed by using these window functions. Windowing methods are used for signal local vertex spectrum calculation with a test signal. Norm one based Concentration Measure is used for comparison.

  • Two-component bivariate signal decomposition based on time-frequency analysis
    2017 22nd International Conference on Digital Signal Processing (DSP), 2017
    Co-Authors: Ljubiša Stanković, Miloš Daković, Miloš Brajović, Danilo Mandić
    Abstract:

    A time-frequency analysis based approach for the decomposition of bivariate signals is presented. In particular, the well-known problem of two components overlapping in the time-frequency plane while having non-linear instantaneous frequencies is considered. The bivariate form of data leads to a significant modification of the Wigner distribution cross-terms. Therefore, the eigenvalue decomposition of Wigner distribution based signal autocorrelation matrix produces two significant eigenvalues instead of one in the common Wigner distribution. It is shown that the two corresponding eigenvectors can be linearly combined in order to produce fully separated signal components. The unknown coefficients are found by minimizing the time-frequency Concentration Measure of these particular eigenvectors linear combination. The presented approach is illustrated on the decomposition of a fast-varying real-valued signal with small instantaneous frequencies, so that its positive and negative frequency parts are so close that they degrade the analytical signal representation.

  • Adaptive local polynomial Fourier transform
    2002 11th European Signal Processing Conference, 2002
    Co-Authors: Miloš Daković, Igor Djurović, Ljubiša Stanković
    Abstract:

    Adaptive local polynomial Fourier transform (ALPFT) is proposed in this paper. For multicomponent FM signals with parallel instantaneous frequencies (IF) the ALPFT is chosen as the local polynomial Fourier transforms (LPFT) which produces maximal Concentration Measure from a set of the LPFTs. For other forms of the FM signals the ALPFT is determined as a weighted sum of the LPFTs. The weighting coefficients are determined as a function of the Concentration Measure. The proposed transforms produce highly concentrated time-frequency signal representations.

Miloš Daković - One of the best experts on this subject based on the ideXlab platform.

  • Micro-Doppler removal in radar imaging in the case of non-compensated rigid body acceleration
    2018 23rd International Scientific-Professional Conference on Information Technology (IT), 2018
    Co-Authors: Miloš Brajović, Ljubiša Stanković, Miloš Daković
    Abstract:

    The micro-Doppler (m-D) effect, commonly caused by fast moving reflectors, can significantly decrease the readability of rigid body in ISAR/SAR radar images. We revisit an L-statistics based micro-Doppler removal approach, producing excellent results in separation of the stationary rigid body from the m-D. In the case of non-compensated target acceleration, rigid body components become non-stationary, commonly with linear frequency modulation. The Local Polynomial Fourier Transform (LPFT) can be exploited for the estimation of the unknown chirp-rate needed for the acceleration compensation. To this aim, we present a simple iterative procedure based on the LPFT Concentration Measure. It is an alternative to the direct search approach for the estimation of the LPFT demodulation parameter, improving the estimation accuracy and reducing the numerical complexity of the approach. Numerical examples verify the presented theory.

  • Windowing methods for graph signal localization
    2017 22nd International Conference on Digital Signal Processing (DSP), 2017
    Co-Authors: Miloš Daković, Ljubiša Stanković, Ervin Sejdić
    Abstract:

    In this paper we considered windows used for local vertex spectrum analysis of graph signals. In addition to a review of the convolution based windowing method, two methods based on the vertex neighborhood are presented. They are based on the graph path lengths. In the first one the number of edges in a path determine window size, while the edge weights are taken into account in the second method. Signal localization is performed by using these window functions. Windowing methods are used for signal local vertex spectrum calculation with a test signal. Norm one based Concentration Measure is used for comparison.

  • Two-component bivariate signal decomposition based on time-frequency analysis
    2017 22nd International Conference on Digital Signal Processing (DSP), 2017
    Co-Authors: Ljubiša Stanković, Miloš Daković, Miloš Brajović, Danilo Mandić
    Abstract:

    A time-frequency analysis based approach for the decomposition of bivariate signals is presented. In particular, the well-known problem of two components overlapping in the time-frequency plane while having non-linear instantaneous frequencies is considered. The bivariate form of data leads to a significant modification of the Wigner distribution cross-terms. Therefore, the eigenvalue decomposition of Wigner distribution based signal autocorrelation matrix produces two significant eigenvalues instead of one in the common Wigner distribution. It is shown that the two corresponding eigenvectors can be linearly combined in order to produce fully separated signal components. The unknown coefficients are found by minimizing the time-frequency Concentration Measure of these particular eigenvectors linear combination. The presented approach is illustrated on the decomposition of a fast-varying real-valued signal with small instantaneous frequencies, so that its positive and negative frequency parts are so close that they degrade the analytical signal representation.

  • Sparse signal recovery based on Concentration Measures and genetic algorithm
    2016 13th Symposium on Neural Networks and Applications (NEUREL), 2016
    Co-Authors: Miloš Brajović, Miloš Daković, Budimir Lutovac, Irena Orović, Srdjan Stanković
    Abstract:

    In this paper genetic algorithm is applied in the reconstruction of signal with missing samples, sparse in a transformation domain. DFT is considered as a domain of sparsity, without loss of generality. The reconstruction is performed as a minimization of the £1-norm based Concentration Measure, with missing samples acting as minimization variables. Parameters of the genetic algorithm are set based on a numerical study, taking into account the nature of the considered minimization problem. The proposed genetic algorithm parameters setup provides an efficient reconstruction of missing samples under the assumption that the standard reconstruction conditions are met.

  • Adaptive local polynomial Fourier transform
    2002 11th European Signal Processing Conference, 2002
    Co-Authors: Miloš Daković, Igor Djurović, Ljubiša Stanković
    Abstract:

    Adaptive local polynomial Fourier transform (ALPFT) is proposed in this paper. For multicomponent FM signals with parallel instantaneous frequencies (IF) the ALPFT is chosen as the local polynomial Fourier transforms (LPFT) which produces maximal Concentration Measure from a set of the LPFTs. For other forms of the FM signals the ALPFT is determined as a weighted sum of the LPFTs. The weighting coefficients are determined as a function of the Concentration Measure. The proposed transforms produce highly concentrated time-frequency signal representations.

Tharaka A. Lamahewa - One of the best experts on this subject based on the ideXlab platform.

  • Signal Concentration on unit sphere: An azimuthally moment weighting approach
    2010 IEEE International Conference on Acoustics Speech and Signal Processing, 2010
    Co-Authors: Rodney A. Kennedy, Tharaka A. Lamahewa
    Abstract:

    Signals defined on the unit sphere cannot be simultaneously concentrated in a spatial region and in the spherical harmonic spectrum. In this paper we develop a unique spatial Concentration Measure, a kth moment azimuthal Measure, for real-valued spectral-limited signals defined on the unit sphere. The optimal functions with the 2nd and 4th minimum azimuthally moment weighting are obtained and simulation results show: 1) the waveforms tend to be even as k increases, 2) have good energy Concentration in the spatial region, and 3) have better decaying tails than spherical Slepian functions, which form the benchmark set of basis functions for spherical filter design.

  • Signal Concentration on unit sphere: A local k-th moment zenithal energy Concentration Measure
    2010 Australian Communications Theory Workshop (AusCTW), 2010
    Co-Authors: Rodney A. Kennedy, Tharaka A. Lamahewa
    Abstract:

    We define a k-th moment zenithal energy Concentration Measure for a polar cap region on the unit sphere. The optimal functions are obtained with the maximal local k-th moment zenithal Measure (LMZM). Compared to the spherical Slepian function in the cap region and the optimal function with the minimal global k-th moment zenithal Measure (GMZM), we show that: 1) as k increases, the tail of the function with LMZM decays more quickly; 2) the main lobe of the optimal function with LMZM moves as the maximum colatitude of the cap increases; while the spherical Slepian function does not change much and the function with GMZM is by definition invariant; and 3) LMZM has no obvious advantage over the spherical Slepian function and the GMZM optimal function in both the energy Concentration and tail decay rate.

  • Band-limited signal Concentration in time-frequency
    2009 3rd International Conference on Signal Processing and Communication Systems, 2009
    Co-Authors: Rodney A. Kennedy, Tharaka A. Lamahewa
    Abstract:

    This paper studies signal Concentration in time-frequency using the general constrained variational method of Franks. We apply this method to formulate the minimum fourth-moment time-duration Measure in the time interval [-T/2, T/2] for band-limited signals. An orthonormal basis of band-limited functions with the minimum fourth-moment time-duration Measure is obtained. Comparing our optimal band-limited signals with Gabor's function with the minimum second-moment time-duration Measure, simulation results show that our function has better properties: 1) faster decaying rate; 2) larger main lobe; and 3) higher energy Concentration Measure.

Igor Djurović - One of the best experts on this subject based on the ideXlab platform.

  • Adaptive local polynomial Fourier transform
    2002 11th European Signal Processing Conference, 2002
    Co-Authors: Miloš Daković, Igor Djurović, Ljubiša Stanković
    Abstract:

    Adaptive local polynomial Fourier transform (ALPFT) is proposed in this paper. For multicomponent FM signals with parallel instantaneous frequencies (IF) the ALPFT is chosen as the local polynomial Fourier transforms (LPFT) which produces maximal Concentration Measure from a set of the LPFTs. For other forms of the FM signals the ALPFT is determined as a weighted sum of the LPFTs. The weighting coefficients are determined as a function of the Concentration Measure. The proposed transforms produce highly concentrated time-frequency signal representations.

Igor Djurovic - One of the best experts on this subject based on the ideXlab platform.

  • Robust time-frequency representation based on the signal normalization and Concentration Measures
    Signal Processing, 2014
    Co-Authors: Igor Djurovic, Ljubisa Stankovic, Marko Simeunovic
    Abstract:

    An efficient procedure for obtaining time-frequency representations under high influence of impulsive noise is proposed in this paper. The procedure uses the fast Fourier transform based algorithm instead of sorting procedures common in the case of various robust time-frequency representations proposed recently. Concentration Measure is used to select a free parameter of the transform.

  • Robust adaptive local polynomial Fourier transform
    IEEE Signal Processing Letters, 2004
    Co-Authors: Igor Djurovic
    Abstract:

    A robust form of the local polynomial Fourier transform (LPFT) is introduced. This transform can produce a highly concentrated time-frequency (TF) representation for signals embedded in an impulse noise. Calculation of the adaptive parameter in the proposed transform is based on the Concentration Measure. A modified form, calculated as a weighted sum of the robust LPFT, is proposed for multicomponent signals.

  • EUSIPCO - Adaptive local polynomial Fourier transform
    2002
    Co-Authors: Milos Dakovic, Igor Djurovic, Ljubisa Stankovic
    Abstract:

    Adaptive local polynomial Fourier transform (ALPFT) is proposed in this paper. For multicomponent FM signals with parallel instantaneous frequencies (IF) the ALPFT is chosen as the local polynomial Fourier transforms (LPFT) which produces maximal Concentration Measure from a set of the LPFTs. For other forms of the FM signals the ALPFT is determined as a weighted sum of the LPFTs. The weighting coefficients are determined as a function of the Concentration Measure. The proposed transforms produce highly concentrated time-frequency signal representations.