Shiyong Li

MODIFIED WAVENUMBER DOMAIN ALGORITHM FOR THREE-DIMENSIONAL MILLIMETER-WAVE IMAGING

Millimeter-wave (MMW) imaging techniques have been used for the detection of concealed weapons and contraband carried on personnel at airports and other secure locations. The combination of frequency-modulated continuous-wave (FMCW) technology and MMW imaging techniques should lead to compact, light-weight, and low-cost systems which are especially suitable for security and detection application. However, the long signal duration time leads to the failure of the conventional stop-and-go approximation of the pulsed system. Therefore, the motion within the signal duration time needs to be taken into account. Analytical three- dimensional (3-D) backscattered signal model, without using the stop-and-go approximation, is developed in this paper. Then, a wavenumber domain algorithm, with motion compensation, is presented. In addition, conventional wavenumber domain methods use Stolt interpolation to obtain uniform wavenumber samples and compute the fast Fourier transform (FFT). This paper uses the 3- D nonuniform fast Fourier transform (NUFFT) instead of the Stolt interpolation and FFT. The NUFFT-based method is much faster than the Stolt interpolation-based method. Finally, point target simulations are performed to verify the algorithm.

A COMPRESSIVE SENSING APPROACH FOR SYNTHETIC APERTURE IMAGING RADIOMETERS

The aperture synthesis technology represents a promising new approach to microwave radiometers for high-resolution observa- tions of the Earth from geostationary orbit. However, the future ap- plication of the new technology may be limited by its large number of antennas, receivers, and correlators. In order to reduce signiflcantly the complexity of the on-board hardware requirements, a novel method based on the recently developed theory of compressive sensing (CS) is proposed in this paper. Due to the fact that the brightness temperature distributions of the Earth have a sparse representation in some proper transform domain | for example, in terms of spatial flnite-difierences or their wavelet coe-cients, we use the CS approach to signiflcantly undersample the visibility function. Thus the number of antennas, re- ceivers, and correlators can be further reduced than those based on the traditional methods that obey the Shannon/Nyquist sampling the- orem. The reconstruction is performed by minimizing the 1 norm of a transformed image. The efiectiveness of the proposed approach is validated by numerical simulations using the image corresponding to AMSU-A over the Paciflc.

Near-field scattering centers estimation using a far-field 3-D ESPRIT type method

A method is presented for estimating the three-dimensional (3-D) scattering centers of an object under spherical wave illumination. A novel approximation of the distance from the antenna to each scattering center is first formulated, which provides an approach for compensating for the spherical wavefront curvature. Additionally, the approximation transforms the near-field problem into virtual far-field problem. A far-field 3-D ESPRIT type algorithm is also presented and applied to estimating the virtual far-field scattering centers. Finally, the near-field scattering centers are obtained by using the compensation relations. Simulation results show the performance improvement of the proposed method compared with the conventional far-field method.

Modified Cylindrical Holographic Algorithm for Three-Dimensional Millimeter-Wave Imaging

Millimeter-wave (MMW) imaging techniques have been developed for the detection of concealed weapons and plastic explosives carried on personnel at major transportation hubs and secure locations. The combination of frequency-modulated continuous-wave (FMCW) technology and MMW imaging techniques leads to wideband, compact, and cost-efiective systems which are especially suitable for security detection. Cylindrical three-dimensional (3-D) imaging technique, with the ability of viewing multiple sides, is an extension of rectilinear 3-D imaging technique only viewing a single side. Due to the relatively long signal sweep time, the conventional stop-and-go approximation of the pulsed systems is not suitable for FMCW systems. Therefore, a 3-D backscattered signal model including the efiects of the continuous motion within the signal duration time is developed for cylindrical imaging systems. Then, a holographic image reconstruction algorithm, with motion compensation, is presented and demonstrated by means of numerical simulations.

Scattering Centers Measurements Using a Modified Matrix Pencil Method

A modified matrix pencil (MP) method is discussed with applications to the scattering centers measurements in step-frequency radar system. Comparison with the Kumaresan-Tufts (KT) method, a parameter estimation method belonging to the polynomial method, is given out, and it is shown the MP method is more robust than the KT method with respect to noise and to the pencil parameter changes. Also the MP method has the computational advantage over the KT method. Results compared with the IFFT method show that the resolution and precision of the MP method are both higher than those of the IFFT method. The MP method is more sensitive to noise than the IFFT method when the signal-to-noise ratio (SNR) is low. But the robustness can be significantly improved by using the mean of the estimates

Raw data compress method of Synthetic Aperture Radar based on compressive sensing

Nowadays, due to the need of high resolution imaging, huge amount of data are needed to be collected base on Nyquist sampling theorem. However, the acquisition platform cannot afford the computation requirement to process on board, so those data must be sent to the ground so that it can be processed. This paper is focused on the compression of the Synthetic Aperture Radar (SAR) raw data to send as less data as we can ease the burden on the system and reduce the time for transmission. In this paper, we compressed the SAR raw data using compressive sensing method, and we train the sparse basis through K-SVD method. First we use the raw data that we collected to train the sparse basis using K-SVD method, when we get the trained sparse basis, we only need to send part of the raw data with the basis to the ground and the raw data can be recovered perfectly. The result of recovered data and imaging results are given. This method can help us to perform further application research of SAR imaging.

Radar Target Signature Processing While Measuring in Near Field

Radar target signature is usually measured in near field using step frequency radar system, and Fourier-based methods are conventionally applied to signature processing, but bias lies between near field and far field. RCS is an important signature, which may be extrapolated by simple signal processing algorithm, some results are shown. The one-dimension downrange profiles are also researched based on a model, which is the foundation of image technology. While two dimension scattering centers distribution is measured, ISAR processing technology is used. We often solve the question by digital signal processing, electromagnetic field methods are also necessary to calculate wave scattering and illuminating, if two approaches are combined, maybe some interesting results can be obtained, which is a trend

An algorithm based on the compressed sensing for near range two dimensional imaging

Planar array two dimensional (2D) imaging is an important technology. It saves scanning time, but needs a large number of antenna elements. In order to save the cost, we hope to use less antennas to get the same (or better) image results. In this paper, we use sparse planar array and reconstruct the image by the two dimensional fast smoothed L0 (2D-SL0) algorithm. Paraxial Green function is used as sensing matrix. Comparisons of the algorithms based on 2D-SL0 algorithm and the matched filter processing (MFP) are demonstrated by means of numerical simulations. It is obvious that the imaging results by the 2D-SL0 algorithm is much clearer. And the focusing performance of the algorithm based on the 2D-SL0 algorithm is very well, even when we use the sparse antenna array.

A modified 2D ESPRIT method for scattering centers measurements

The JSVD is introduced to the 2D ESPRIT method. And the means and standard deviations of the estimation errors are reduced effectively compared to the initial method. The estimated results of the real data show that the modified 2D ESPRIT provides increased resolution and precision over the FFT method.

Scattering Centers Measurements by a 2-D ESPRIT Type Method

A 2-D ESPRIT type method is presented for the scattering centers measurements in step-frequency radar. This method estimates the signal parameter pairs directly unlike the matrix enhancement and matrix pencil (MEMP) method which contains an additional step to pair the parameters related to each dimension. We use a forward-backward prediction data matrix to increase the estimation accuracy and the robustness to noise. Results compared with the Fourier transform method show that the resolution and precision of the 2-D ESPRIT type method are both higher than those of the Fourier method.