Kei Suwa

A Two-Dimensional Bandwidth Extrapolation Technique for Polarimetric Synthetic Aperture Radar Images

The resolution of a synthetic aperture radar (SAR) image, in range and azimuth, is determined by the transmitted bandwidth and the synthetic aperture length, respectively. Various superresolution techniques for improving resolution have been proposed, and we have proposed an algorithm that we call polarimetric bandwidth extrapolation (PBWE). To apply PBWE to a radar image, one needs to first apply PBWE in the range direction and then in the azimuth direction, or vice versa . In this paper, PBWE is further extended to the 2-D case. This extended case (2D-PBWE) utilizes a 2-D polarimetric linear prediction model and expands the spatial frequency bandwidth in range and azimuth directions simultaneously. The performance of the 2D-PBWE is shown through a simulated radar image and a real polarimetric SAR image

Three-Dimensional Target Geometry and Target Motion Estimation Method Using Multistatic ISAR Movies and Its Performance

Inverse synthetic aperture radar (ISAR) is one of the radar techniques used to observe 2-D images of a remotely based target using radio waves. If we keep observing the target and consecutively generate multiple ISAR images, which we call an ISAR movie, the target image varies considerably due to the motion of the target. The authors have proposed an algorithm for reconstructing a 3-D target shape from an ISAR movie; however, the algorithm requires a priori knowledge of the relative motion of the target. In this paper, we propose a novel method that estimates the relative motion and the 3-D shape of the target using a multistatic ISAR movie.

A bandwidth extrapolation technique of polarimetric radar data and a recursive method of polarimetric linear prediction coefficient estimation

Resolution of a radar is limited in range by its bandwidth. One of the various super-resolution techniques for improving its resolution is bandwidth extrapolation (BWE). In the technique, a linear prediction model is fitted to the data, and the model is used to extrapolate the bandwidth. In this paper, the concept of BWE is extended, and a new algorithm called polarimetric bandwidth extrapolation (PBWE) applicable to polarimetric radar data is proposed. It is shown through numerical simulations that utilization of full polarization information allows PBWE to improve the resolution beyond the conventional BWE method. Some results of physical simulation experiment using a W-band polarimetric FMCW radar and corner reflectors are shown to confirm the advantage of PBWE.

A bandwidth extrapolation technique for improved range resolution of polarimetric radar data

Radar resolution is limited in range by its bandwidth. One of the various super-resolution techniques for improving resolution is bandwidth extrapolation (BWE). In this paper, the concept of BWE is extended, and a new algorithm called polarimetric bandwidth extrapolation (PBWE), applicable to polarimetric radar data, is proposed. It is shown through simulations that utilization of full polarization information allows PBWE to improve the resolution beyond the conventional BWE method.

Image based approach for target detection and robust target velocity estimation method for multi-channel SAR-GMTI

This paper presents new algorithms for moving target detection and velocity estimation for multi-channel SAR (Synthetic Aperture Radar) system. The algorithms are derived as the extensions of conventional DPCA (Displaced Phase Center Antenna) and ATI (Along Track Interferometry) to the multichannel case. The proposed multi-channel DPCA based on orthogonal projection completely suppresses the static clutter even with the presence of the azimuth ambiguity. The multi-channel ATI utilizes the phase differences of all the pairs of the channels to estimate the target radial velocity. The combination of the two, which we call multi-channel DPCA-ATI, is also proposed as the more robust velocity estimation method.

An experiment of azimuth ambiguity suppression by multiple receiver apertures with airborne Ku-band synthetic aperture radar

This paper presents an experimental result to demonstrate azimuth ambiguity suppression with the displaced phase center antenna technique that is promising technique to realize the high resolution wide swath imagery for spaceborne synthetic aperture radar (SAR). The dual receiver antenna data was acquired by airborne Ku-band SAR. In this paper, we propose a signal compensation procedure to cope with beam squint for the following coherent signal processing such as the dual channel reconstruction algorithm. The azimuth ambiguity suppression capability is evaluated by comparing the image reconstructed from dual-receiver channel to the single receiver channel image. The experimental result shows significant azimuth ambiguity suppression.

Forward looking radar imaging method using multiple receiver antennas and digital beam forming technique

Synthetic Aperture Radar (SAR) suffers from two major difficulties when it is applied to forward looking direction. One is that the achievable cross range resolution becomes low due to the small gradient of the Doppler frequency around the flight direction. The other is the left / right ambiguity, which is caused by the fact that the targets situated symmetrically around the flight path have the same Doppler information. In this paper, a forward looking SAR method which utilizes multiple receive antennas is proposed. The proposed method first forms multiple beam patterns that suppress the symmetric directions by applying the digital beam forming technique, and the SAR image is formed with the appropriate choice of the beams. It is shown through numerical simulations that the utilization of multiple antennas and beamforming technique eliminates the left / right ambiguity.

A simulator of synthetic aperture radar images; land, ocean surface, and man-made targets

Synthetic aperture radar (SAR) images are very useful for many applications. However, SAR images require a great deal of experience for the interpretation, since the appearances are very different from optical images. Additionally, the appearances change significantly by the observational condition, such as frequency and incident angle. We propose a SAR image simulating algorithm developed for the aid of the SAR image analysis and training. The simulator is also intended to support the SAR system configuration design by providing the expected images. The main features of the algorithm are the capability of wide area simulation and the low computational costs. The simulated image of 10 km square area is shown and compared with the real image

Analysis on the Resolution of Polarimetric Radar and Performance Evaluation of the Polarimetric Bandwidth Extrapolation Method

Polarimetric radar provides useful information on scattering mechanisms, and advances in polarimetric synthetic aperture radar (POLSAR) technology have shown the effect of polarization information on various applications. The disadvantage of POLSAR is low azimuth resolution capability. Since the effective pulse repetition frequency (PRF) in each polarimetric channel of POLSAR is half the total PRF, the resolution is twice as low as the single-polarization SAR. This drawback may be mitigated by employing a superresolution method designed for polarimetric radar such as polarimetric bandwidth extrapolation (PBWE), previously proposed by the authors. Thus far, the authors only empirically showed that polarization information helps in improving resolution via numerical simulations. In this paper, theoretical analysis on the resolution of a polarimetric radar is provided. Along with the resolution, estimation accuracy of the polarization property of the target is studied. In the analysis, we employ the recently proposed metric called the statistical resolution limit (SRL) that provides the highest resolution achievable by any unbiased parametric spectral estimator. The SRL of polarimetric radar is derived and compared with that of single-polarization radar. The Cramér-Rao bound (CRB) of the polarization estimation error of the polarimetric radar is also derived. The complexified full Fisher information matrix and the CRBs of signal parameters are derived first for the general multichannel signal in nonwhite Gaussian noise. Then, the SRL and the polarization estimation error are derived using the CRBs of signal parameters for the cases where two point targets are closely located. It is shown analytically and numerically that the polarization information helps in improving the SRL. It is also shown that the polarimetric processing significantly improves the accuracy of the polarization estimation when two targets are located closer than the Fourier resolution cell. Finally, the performance of the PBWE is evaluated via simulations and is compared to the SRL and the CRB for the polarization estimation.

Transmission data reduction in bandwidth scarce environment for the earth surface observation from satellites flying in close formation

The remote sensing Synthetic using Aperture Radar can obtain the information of moving targets on the earth surface using MTI (Moving Target Indication) technique, which detects differentiation of the surface image data obtained in successive sessions taken with two antennas mounted on one satellite or two satellites flying in a close formation. In the latter case, the operation of the differentiation is usually performed at the ground station, then each satellite needs to transmit the large amount of data to the ground station separately, which requires broad downlink bandwidths. However, the earth surface image data at the two satellites clearly have a strong correlation since the static background image data other than the moving targets are essentially the same except for some noise factors. Focusing on this fact, we claim that the transmission data can and should be reduced by taking off the redundancy of the two data and sending only the differential parts. In this paper, we discuss the coding theory matter related to this redundancy suppression first, and then discuss source, channel and network coding, and also the Slepian-Wolf theorem for two correlated sources. We next propose a coding and compressing scheme for a certain simplified model. We further generalize it in order to apply this idea to a more realistic model and present a procedure of the proposed scheme.