This letter presents a clutter statistical analysis for stacks of wavelength-resolution synthetic aperture radar (SAR) images. Each image stack consists of SAR images generated by the same sensor, using the same flight track illuminating the same scene but with a time separation between the illuminations. We test three candidate statistical distributions for time changes in the stack, namely, Rician, Rayleigh, and log-normal. The tests results reveal that the Rician distribution is a very good candidate for modeling stack of wavelength-resolution SAR images, where 98.59 & x0025; of the tested samples passed the Anderson-Darling (AD) goodness-of-fit test. Also, it is observed that the presence of changes in the ground scene is related to the tested samples that have failed in the AD test for the Rician distribution hypothesis.
Radio Occultation based on Global Navigation Satellite System signals (GNSS RO) is an increasingly important remote sensing technique. Its measurements are used to derive parameter of the Earth's atmosphere, e.g., pressure, temperature and humidity, with good accuracy. The systematic residual error present on the data processing is related to ionospheric conditions, such as the distribution of electrons and the resultant vertical gradient. This study investigates the relationship between these parameters and the residual ionospheric error (RIE) on the retrieved bending angle in the stratosphere. Chapman function combined to sinusoidal perturbations are used to model electron density profiles and compared to RO retrievals of the ionosphere to perform the investigation. The results confirmed that the major ionospheric influence on the retrieval error is related to the F-layer electron density peak, whereas small-scale vertical structures play a minor role.
The aim of this letter is to compare two incoherent change-detection algorithms for target detection in low-frequency ultrawideband (UWB) synthetic aperture radar (SAR) images. The considered UWB SAR operates in the frequency range from 20 to 90 MHz. Both approaches employ a likelihood ratio test according to the Neyman–Pearson criterion. First, the bivariate Rayleigh probability distribution is used to implement the likelihood ratio test function. This distribution is well known and has been used for change-detection algorithms in low-frequency UWB SAR with good results. Aiming to minimize the false alarm rate and taking into consideration that low-frequency UWB SAR images have high resolution compared to the transmitted wavelength, the second approach implements the test by using a bivariate K-distribution. This distribution has scale and shape parameters that can be used to adjust it to the data. No ﬁlter is applied to the data set images, and the results show that with a good statistical model, it is not needed to rely on ﬁltering the data to decrease the number of false alarms. Therefore, we can have a better tradeoff between resolution and detection performance.
In this paper we propose a new change detection (CD) algorithm based on the Bayes theorem and probability assignments. Differently from any kind of likelihood ratio test (LRT) algorithms, the proposed algorithm does not present target alarms, but the probability of certain image position is a target position. In other words, the proposed method leads to quantitative estimates on the probability of a target at any pixel, whereas LRT algorithms can only be used as a figure of merit for any pixel to contain a target.
Wave optics propagators (WOPs) are commonlyused to describe the propagation of radio signals through earth’satmosphere. In radio occultation (RO) context, multiple phasescreen (MPS) method has been used to model the effects of theatmosphere in Global Navigation Satellite System (GNSS) signalsduring an occultation event. WOP implementation includes,in addition to MPS, a diffraction integral as the final step tocalculate the radio signal measured in the low-earth orbit (LEO)satellite. This approach considers vacuum as the propagationmedium at high altitudes, which is not always the case when theionosphere is taken into account in simulations. An alternativeapproach is using MPS all the way to LEO in order to samplethe GNSS signal in orbit. This approach, named MPS orbitsampling (MPS-OS), is evaluated in this letter. Different scenariosof setting occultation assuming a short segment of the LEO orbithave been simulated using MPS and MPS-OS. Results have beencompared to Abel transform references. Furthermore, a longsegment scenario has been evaluated as well. A comparison ofbending angle (BA) and residual ionospheric error (RIE) showsthe equivalence between MPS and MPS-OS results. The mainapplication of MPS-OS should be in occultation events with longsegments of orbit and including ionosphere, in which a standardWOP may not be appropriate.
Like any other system relying on trans-ionospheric propagation, GNSS Radio Occultation (GNSS-RO) is affected by ionospheric conditions during measurements. Regions of plasma irregularities in F-region create abrupt gradients in the distribution of ionized particles. Radio signals propagated through such regions suffer from constructive and destructive contributions in phase and amplitude, known as scintillations. Different approaches have been proposed in order to model and reproduce the wave propagation through ionospheric irregularities. We present simulations considering an one-component inverse power-law model of irregularities integrated with Multiple Phase Screen (MPS) propagation. In this work, the capability of the scintillation model to reproduce features in the signal amplitude of low latitude MetOp measurements in the early hours of DOY 76, 2015 (St. Patrick’s Day geomagnetic storm) is evaluated. Power spectral density (PSD) analysis, scintillation index, decorrelation time and standard deviation of neutral bending angle are considered in the comparison between the simulations and RO measurements. The results validate the capability of the simulator to replicate an equivalent total integrated phase variance in cases of moderate to strong scintillation.
This paper presents an analysis of pre-filtered clutter VHF SAR images. The image data are reorganized into sub-vectors based on the observation of the image-pair magnitude samples. Based on this approach, we present a statistical description of the SAR clutter obtained by the subtraction between two real SAR images. The statistical analysis based on bivariate distribution data organized into different intervals of magnitude can be an important tool to further understand the properties of the backscattered signal, which can be a valuable premise for change detection processing.
Change detection is an important synthetic aperture radar (SAR) application, usually used to detect changes on the ground scene measurements in different moments in time. Traditionally, change detection algorithm (CDA) is mainly designed for two synthetic aperture radar (SAR) images retrieved at different instants. However, more images can be used to improve the algorithms performance, witch emerges as a research topic on SAR change detection. Image stack information can be treated as a data series over time and can be modeled by autoregressive (AR) models. Thus, we present some initial findings on SAR change detection based on image stack considering AR models. Applying AR model for each pixel position in the image stack, we obtained an estimated image of the ground scene which can be used as a reference image for CDA. The experimental results reveal that ground scene estimates by the AR models is accurate and can be used for change detection applications. © 2018 SPIE.
One important application of Synthetic Aperture Radars (SAR) is positioning of targets with high accuracy in both azimuth and range. If the target is moving and a multi-channel SAR system is used also the speed components in azimuth and range can be found with a high accuracy. In this paper we propose a method to estimate the accuracy of such a multichannel SAR system. The method is based on the Cramér-Rao Lower Bound (CRLB). To exemplify the method the variance of parameter estimates by a single channel UHF UWB SAR system is found.
The paper introduces a new likelihood ratio test (LRT) for incoherent detection of man-made objects obscured by foliage in forest area. The test is performed to detect changes between a reference image and a surveillance image. The method is developed for change detection in high resolution Synthetic Aperture Radar (SAR). For simplicity and lack of more appropriate models, the new LRT is still based on simple and efficient models. If there is no man-made object, the statistical model for clutter and noise of two images will be a bivariate Rayleigh distribution. In contrary, a joint distribution of Rayleigh and uniform is used to model for target, clutter, and noise. The proposed LRT is evaluated using radar data acquired by CARABAS in northern Sweden. The probability of detection is up to 96% with much less than one false alarm per square kilometer. © 2017 IEEE.
This paper addresses multi-dimensional Ground Moving Target Indication (GMTI) using a multi-channel Wide Band (WB) Synthetic Aperture Radar (SAR) system. For limited time intervals the target acceleration is so small that target motion can be related to two ground and two speed coordinates. However, four other dimensions are used in WB SAR GMTI processing during the detection phase: azimuth, range, bearing, and the relative speed between the object and the SAR platform. In the detection phase, blind hypotheses are used, and the discretization steps between the hypotheses are a trade-off between the number of hypotheses tested and detectability. As the integration angle increases, the bandwidth increases and the therefore the number of tests increases. In this paper we discuss the discretization step in all four dimensions for moving target detection, and analyze the step size in particular in the most critical domain, the relative speed. The analysis is made on CARABAS II data.
This letter presents an analysis of prefiltered clutter ultrawideband (UWB) very high frequency synthetic aperture radar (SAR) images. The image data are reorganized into subvectors based on the observation of the image-pair magnitude samples. Based on this approach, we present a statistical description of the SAR clutter obtained by the subtraction between two real SAR images. The statistical analysis based on bivariate distribution data organized into different intervals of magnitude can be an important tool to further understand the properties of the backscattered signal for low-frequency SAR images. In this letter, it is found that, for “good” image pairs, the subtracted image has Gaussian distributed clutter backscattering and that the noise mainly consists of the thermal noise and, therefore, speckle noise does not have to be considered. This is a consequence of the stable backscattering for a UWB low-frequency SAR system.
High accuracy of impact height is important to get reliableRadio Occultation (RO) measurements of the atmosphere refractivity.We have made an investigation on how accuratelywe can measure the impact height at ground level using waveoptics simulations, realistic refractivity profiles, a realisticsimulator for an advanced RO instrument including noise,and using phase matching for the inversion. The idea of theinvestigation is to increase the measurement accuracy of impactheight at low altitudes and to give reliable measurementseven in cases of super-refractive layers. We present statisticson the accuracy and precision of the determination of theimpact height at ground, as well as the resulting accuracy andprecision in the measured refractivity.
Global Navigation Satellite System Radio Occultation (GNSS-RO) is an important technique used to sound the Earth's atmosphere and provide data products to numerical weather prediction (NWP) systems as well as toclimate research. It provides a high vertical resolution and SI-traceability that are both valuable complements toother Earth observation systems. In addition to direct components refracted in the atmosphere, many received RO signals contain reflected components thanks to the specular and relatively smooth characteristics of the ocean. These reflected components can interfere the retrieval of the direct part of the signal, and can also contain meteorological information of their own, e.g., information about the refractivity at the Earth's surface. While the conventional method to detect such reflections is by using radio-holographic methods, it has been shown that it is possible to see reflections using wave optics inversion, specically while inspecting the amplitude of the output of phase matching (PM). The primary objective of this paper is to analyze the appearance of these reflections in the amplitude output from another wave optics algorithm, namely the much faster full spectrum inversion (FSI). PM and FSI are closely related algorithms - they both use the method of stationary phase to derive the bending angle from a measured signal. We apply our own implementation of FSI to the same GNSS-RO measurements that PM was previously applied to and show that the amplitudes of the outputs again indicate reflection in the surface of the ocean. Our results show that the amplitudes output from the FSI and PM algorithms are practically identical and that the reflection signatures thus appear equally well.
TerraSAR-X and TanDEM-X has since 2014 been flying in the monostatic pursuit mode. This mode consists of a formation flying with the satellites having identic orbit, but with a displacement in along-track of a distance corresponding to 10 s delay. Such formation gives e.g. the ability to detect movements of very slow targets or targets standing still and moving between measurements. Depending on the integration time for each satellite and the time separation between the satellites, the clutter may behave as stationary or non-stationary. This allows for the possibility to apply either coherent or incoherent change detection algorithms. As has been proved earlier, the potential of coherent change detection has very good abilities but has proven difficult to obtain on X-band. In the case of very high resolution SAR, there will also be a target smearing effect due to the target movement within the coherent processing interval (CPI). The target defocusing effect may also be used for detection of moving targets in the scene using the so-called Detection of Moving Targets by Focusing (DMTF) technique. This is since an estimate or hypothesis test for target movements can be used to reprocess the SAR image and obtain a Signal-To-Clutter-Noise (SCNR) gain, thus increase chances for moving target detection. Therefore, a combination of DMTF and change detection over short times is promising. As such, this paper investigates the potential of the monostatic pursuit mode for GMTI.
In this paper, results of moving target detection in multichannel UHF-band Synthetic Aperture Radar (SAR) data are shown. The clutter suppression is done using Finite Impulse Response (FIR) filtering of multichannel SAR in combination with a 2-stage Fast Backprojection (FBP) algorithm to focus the moving target using relative speed. The FIR filter coefficients are chosen with the use of STAP filtering. Two parameters are used for target focusing, target speed in range and in azimuth. When the target is focused, both speed parameters of the target are found. In the experimental results, two channels were used in order to suppress clutter. In the resulting SAR images it is obvious that very strong scatterers and the forest areas have been suppressed in comparison to the moving target in the image scene. The gain obtained can be measured using SCNR gain, which is about 19dB. Another way to measure signal processing gain is the ability to suppress the strongest reflecting object in the SAR scene. The gain of target in relation to this object is 25dB. This shows that using UHF-band SAR GMTI for suppressing forest and increasing the target signal can work.
In this paper, an investigation is made on how sidelobes can be suppressed in ultrawideband-ultrawidebeam (UWB) Synthetic Aperture Radar (SAR) using apodization. Due to the special properties of UWB SAR such as very wide integration angle and very large relative bandwidth, the support for the spectrum of a SAR image differs distinctively from a rectangle, which is the normal approximation in narrowband-narrowbeam (NB) SAR. The proposal in the paper is to apply non-separable windows to the spectrum, in order to suppress sidelobes. Non-separable windows are shown to give less broadening of mainlobe while maintaining the same suppression of sidelobes in comparison to separable windows. In the comparison, parameters for three different SAR systems are used.
This paper presents a comparative study of the polar and the subimage based variants of the time domain SAR algorithm Fast Factorized Backprojection. The difference between the two variants with regard to the phase error, which causes defocusing in the image, is investigated. The difference between the algorithms in interpolation between stages is also discussed. To investigate the sidelobes in azimuth, the paper gives simulation results for a low frequency UWB SAR system for both algorithms. How the algorithms differ with regard to amount of beams and length of beams is also discussed.
In the area of SAR imaging, it is of interest to be able to focus moving targets. In this paper, an algorithm for moving target focusing is presented. The algorithm is able to refocus a smeared moving target in a SAR image processed at one relative speed to the correct one.. The algorithm works in the frequency domain and is based on the Range Migration algorithm. The refocusing can be made on the whole SAR image or small sub images corresponding to physical areas of interest for the end user. By applying the algorithm to a small image, the computational cost is greatly reduced compared with using the full SAR image. The performance is illustrated by applying the algorithm to simulated SAR data according to the parameters for the LORA system.
This paper presents an iterative method to estimate the Normalised Relative Speed (NRS) of ground moving targets in Ultra Wideband (UWB) wide beam Synthetic Aperture Radar (SAR) using one antenna. The number of iterations depends on the separation between processed NRS and true target NRS. The NRS estimate is based on a chirp rate estimator in azimuth direction of the SAR image. The paper derives an analytical expression of the azimuth phase information based on the moving target NRS and the NRS used in the image formation. The method has been tested on real data from the CARABAS-II SAR system showing good results.
In this paper, a method for moving target relative speed estimation and refocusing based on synthetic aperture radar (SAR) images is derived and tested in simulation and on real data with good results. Furthermore, an approach on how to combine the estimation method with the refocusing method is introduced. The estimation is based on a chirp estimator that operates in the SAR image and the refocusing of the moving target is performed locally using subimages. Focusing of the moving target is achieved in the frequency domain by phase compensation, and therefore makes it even possible to handle large range cell migration in the SAR subimages. The proposed approach is tested in a simulation and also on real ultrawideband (UWB) SAR data with very good results. The estimation method works especially well in connection with low frequency (LF) UWB SAR, where the clutter is well focused and the phase of the smeared moving target signal becomes less distorted. The main limitation of the approach is target accelerations where the distortion increases with the integration time.
TerraSAR-X and TanDEM-X is able to fly in monostatic pursuit mode, a formation with the satellites in identic orbit displaced in along-Track of a distance corresponding to 10 s delay. Such formation gives e.g. The ability to detect very slow targets or targets moving only between measurements. This can be performed using Change Detection (CD) and/or the Detection of Moving Targets by Focusing (DMTF) technique. A combination of DMTF and CD is promising. To investigate the applicability of the methods for moving target detection, an experiment was performed with TerraSAR-X and TanDEM-X in monostatic pursuit mode and several deployed targets. © VDE VERLAG GMBH · Berlin · Offenbach.
The paper introduces a simple experimental groundbased SAR system for studying SAR fundamentals. The SAR system is developed on a vector network analyzer (VNA), for example Agilent E5071C, with some useful built-in functions such as transform and gating. The procedure of acquiring the data by using the SAR system is presented in details. The acquired empirical data is also used to reconstruct the illuminated scene. The possibilities to use the SAR system to support SAR research topics are also discussed in this paper.
The paper presents another possibility to focus moving targets using normalized relative speed (NRS). Similar to the currently used focusing approach, the focusing approach proposed in this paper aims at the ultrawideband and ultrawidebeam synthetic aperture radar systems (UWB SAR) like CARABAS-II. The proposal is shown to overcome the shortcomings of the original focusing approach and can be extended to more complicated cases, for example bistatic SAR.
The paper presents a study of the capability of time- And frequency-domain algorithms for bistatic SAR processing. Two typical algorithms, Bistatic Fast Backprojection (BiFBP) and Bistatic Range Doppler (BiRDA), which are both available for general bistatic geometry, are selected as the examples of time- And frequency-domain algorithms in this study. Their capability is evaluated based on some criteria such as processing time required by the algorithms to reconstruct SAR images from bistatic SAR data and the quality assessments of those SAR images.
A 2-D spectrum for bistatic synthetic aperture radar is derived in this letter. The derivation is based on the commonly used mathematic principles such as themethod of stationary phase and the Fourier transform and the Lagrange inversion theorem in order to find the point of stationary phase in the method of stationary phase. Using the Lagrange inversion theorem allows minimizing the initial assumptions or the initial approximations. The derived 2-D spectrum is compared with the commonly used 2-D spectrum to verify it and illustrate its accuracy.
The paper presents investigations on SAR image statistics and adaptive signal processing for change detection. The investigations show that the amplitude distributions of SAR images with possibly detected changes, that is retrieved with a linear subtraction operator, can approximately be represented by the probability density function of the Gaussian or normal distribution. This allows emerging the idea to use the available adaptive signal processing techniques for change detection. The experiments indicate the promising change detection results obtained with an adaptive line enhancer, one of the adaptive signal processing technique. The experiments are conducted on the data collected by CARABAS, a UWB low frequency SAR system.
Ground moving target indication (GMTI) is one of the important applications of synthetic aperture radar (SAR). This paper introduces a GMTI method for local detection of the ground moving targets in SAR image based on normalized relative speed (NRS) hypotheses. The input of the GMTI method is a complex SAR image where areas of interest for GMTI are locally selected. The output of the method includes the detected targets, the NRSs with respect to the SAR platform, the locations where the detected targets are focused, and the SAR images of the detected targets. The mathematical background of the GMTI method is presented in detail. The method is validated with the data delivered by the Coherent All-Radio Band Sensor, an airborne ultrawideband-ultrawidebeam low-frequency SAR system. The shortcomings of the introduced method are investigated and followed with the solutions. IEEE
The practical considerations in ultrawideband (UWB) synthetic aperture radar (SAR) data processing in general and UWB SAR imaging in particular are clarified and presented in detail in this thesis. They are imaging algorithm, impulse response function in SAR imaging (IRF-SAR), apodization, RF interference (RFI) and SAR image quality measurement. Different algorithms in both time- and frequency domain and their suitability to process UWB SAR data are investigated and evaluated. The necessary modifications for these algorithms are proposed to fulfill the requirements of UWB SAR data processing. The time-domain imaging algorithms are highly recommended for UWB SAR data processing due to their characteristics such as integrated motion error compensation, unlimited scene size and local processing. A new IRF-SAR, which is a function of fractional bandwidth and antenna beamwidth, is derived. The function allows us to investigate different UWB SAR systems. Such investigations are not facilitated by currently used IRF-SAR, Sinc functions. The derived IRF-SAR is totally valid to investigate narrowband (NB) SAR systems. A discussion about the apodization techniques and possibilities to apply to UWB SAR data processing is given in this thesis. Handling orthogonal and non-orthogonal sidelobe in UWB SAR imaging is shown to be challenging with the currently used apodization approaches. The linear apodization approaches always result in the loss in resolutions while the phase information can be destroyed by the nonlinear apodization approaches. A new approach to suppress RFI in UWB SAR signal, which is easy to be disturbed by RFI sources, is suggested. The advantages of the approach compared to the others can be found in adaptive and real time processing characteristics. A new definition of SAR image quality measurement is also presented in this thesis. The complicated behavior of IRF-SAR over fractional bandwidth and antenna beamwidth results in the unsuitability of the currently used definition for UWB SAR image quality measurements. The unsuitability is mainly caused by the inappropriate delimitation of mainlobe and sidelobe areas, the fixed broadening factors and the fixed spreading factor of the orthogonal and non-orthogonal sidelobes. Based on these practical considerations, the thesis also presents some possibilities to propose a definition of UWB SAR which is still not available. The beginning investigated results show that these possibilities comply with the UWB definition proposed by Federal Communications Commission (FCC) in 2002.
This dissertation presents practical issues in Ultrawideband – Ultrawidebeam (UWB) Synthetic Aperture Radar (SAR) signal processing and crucial applications developed on UWB SAR. In the context of this dissertation, UWB SAR refers to the SAR systems utilizing large fractional bandwidth signals and synthesizing long apertures associated with wide antenna beamwidths. On one hand, such specific systems give us opportunities to develop unique applications. One the other hand, signal processing for data collected by these systems is very challenging and therefore requires much effort due to their characteristics. In the signal processing part, the tools supporting the UWB SAR system design and evaluation are introduced. They include an Impulse Response Function in UWB SAR imaging (IRF-SAR), azimuth and range resolution equations for UWB SAR, and a definition of UWB SAR quality measurements. Pre-processing, processing and post-processing for UWB SAR are also topics that will be examined in the signal processing part. The processing is here defined by SAR algorithms. With this definition, the pre-processing refers to RFI suppression approaches whereas the post-processing implies apodization or sidelobe control methods. In the application part, Ground Moving Target Indication (GMTI) is selected for study due to its interest to both military and civilian end-users. GMTI developed on UWB SAR relates to the moving target detection by focusing technique which can be combined with the space-time processing such as Displaced Phase Center Antenna (DPCA) and Space-Time Adaptive Processing (STAP).
This paper introduces an effective detector for wavelength-resolution SAR incoherent change detection. The detector is derived from Bayes' theorem. The input of the detector is the differences between surveillance and reference magnitude images simply obtained by a subtraction while the output is a summary of the detected changes. The proposed detector is tested with 24 CARABAS images that were obtained from the measurement campaign in northern Sweden in 2002. The testing results show that the detector can provide a high average detection probability, e.g., about 96%, with a very low false alarm rate, e.g., only 0.35 per square kilometer. © 2019 IEEE.
A gamma probability density function (pdf) is shown to be an alternative to model the distribution of the magnitudes of high-resolution, i.e., wavelength-resolution, synthetic aperture radar (SAR) images. As investigated in this paper, it is more appropriate and more realistic statistical in comparison with, e.g., Rayleigh. A bivariate gamma pdf is considered for developing a statistical hypothesis test for wavelength-resolution incoherent SAR change detection. The practical issues in implementation of statistical hypothesis test, such as assumptions on target magnitudes, estimations for scale and shape parameters, and implementation of modified Bessel function, are addressed. This paper also proposes a simple processing scheme for incoherent change detection to validate the proposed statistical hypothesis test. The proposal was experimented with 24 CARABAS data sets. With an average detection probability of 96%, the false alarm rate is only 0.47 per square kilometer. IEEE
SAR systems synthesizing circular apertures have been shown to result in better spatial resolutions than the ones synthesizing linear apertures. The paper presents an investigation about the enhancement of SAR spatial resolutions with the use of circular aperture. A comparison between the spatial resolutions obtained with a SAR system synthesizing a circular aperture and with the same SAR system synthesizing a linear aperture is therefore carried out. The studying results are verified by the experimental SAR data set provided by the experimental ground-based SAR system of Blekinge Institute of Technology (BTH GB-SAR).
This letter introduces ground-range and cross-range resolution equations for the side-looking bistatic synthetic aperture radar (SAR). The derivation is based on the backprojection integral and the method of stationary phase. The ground-range and cross-range resolution equations are provided in closed form, making them easy for calculation. They are, therefore, helpful for bistatic SAR system development. The derived ground-range and cross-range resolution equations are validated with the bistatic data simulated mainly using the parameters of the LORA system. IEEE
Left/right ambiguity and low Doppler resolution are severe problems for monostatic forward-looking SAR imaging. Among approaches for the problems, using bistatic SAR is very promising. The left/right ambiguity in bistatic forwardlooking SAR has been investigated in detail recently whereas an appropriate study for enhancing bistatic forward-looking SAR Doppler resolution is still desired. A research on Doppler resolution and then cross-range resolution in bistatic forward-looking SAR is therefore presented in this paper. The research is based on the use of the gradient. Several recommendations which allow maximizing Doppler resolution and cross-range in bistatic forward-looking SAR imaging are provided.
This paper introduces a bistatic fast backprojection synthetic aperture radar (SAR) algorithm that is available for different bistatic geometries. The proposed algorithm is tested with the bistatic CARABAS-like data, and the results indicate that the algorithm is a good candidate for bistatic SAR data processing. The image quality measurements are quite similar to the referenced values obtained with the bistatic global backprojection algorithm. That is, the peak sidelobe ratio (PSLR) is -13.7 dB in comparison to the referenced PSLR of -13.8 dB. The half-power beamwidths (HPBWs) measured on the cuts in x and y and the direction where the peak sidelobes locate are 2.8, 3.7, and 3.5 m, respectively, which are approximate to the referenced HPBWs. The small differences in the measured results mainly come from the interpolation step of the algorithm.
The paper proposes a ground moving target detection and estimation method aiming at Ultra Wide Band and -Beam (UWB) Synthetic Aperture Radar (SAR) systems. The method is developed on the moving target detection by focusing technique and requires a SAR system flying with two different speeds during the integration time. The method allows us to detect ground moving target, even hidden by clutter, and to estimate the target parameters, i.e. speed and direction of motion. The accuracy of the estimations depends strongly on the computational cost and can therefore be controlled. The proposal is tested with the simulated CARABAS data.
The paper presents a derivation of Nyquist sampling requirements for the polar grids in some bistatic time-domain algorithms. The derivation is based on an airborne bistatic system with general bistatic geometry. The Nyquist sampling requirements are shown to be the functions of operating radar frequency, transmitter and receiver subaperture lengths, and bistatic geometry. How to decide the Nyquist sampling requirements for different bistatic geometries and the relationship between the Nyquist sampling requirements in the monostatic and bistatic cases are also addressed in the paper. The derived Nyquist sampling requirements is examined with the bistatic CARABAS-II like data.
Left/right ambiguity and low angular (azimuth) resolution are severe problems for monostatic forward-looking SAR imaging. It is strongly believed that these technical issues can definitely be solved with bistatic forward-looking SAR. The analysis presented in this paper points out that the left/right ambiguity problem still exits. However, an appropriate selection of the position of bistatic base line and antenna beamwidth allows us to conceal it. The paper also gives some recommendations which can be considred for the forward-looking SAR imaging.
The paper introduces a Range Migration algorithm for bistatic SAR data processing. The algorithm is developed on the available function representing the two-dimension Fourier transform of the bistatic SAR data and relationship between radar signal frequency and wave-numbers for bistatic SAR or the ω - k relationship. The algorithm is tested with the simulations using different SAR geometries for validation. Some inherit limits of the developed Range Migration algorithm and the possible solutions are also discussed in the paper. © 2018 IEEE.
This paper presents a method to implement SAR slow-time space time adaptive processing (STAP) in the beamforming stage of the fast backprojection algorithm. This method is different from the recently published method where the SAR fast-time STAP is implemented after the beamforming stage for detection and then imaging. As a common SAR STAP method, the method proposed in this paper can be used for ground moving target indication (GMTI) and reconstruction of the image of the detected moving targets. The paper also presents some simulation results in order to illustrate the proposed method.
One-dimensional (1-D) adaptive noise canceler (ANC) has been used for false alarm reduction in low frequency SAR change detection. The paper presents possibilities to process two-dimensional (2-D) data by an 1-D ANC. Beside concatenating the rows of 2-D data in a matrix form to convert it to 1-D data in a vector form, two conversion approaches are considered: concatenating the columns of 2-D data and local concatenation, i.e., the conversion to 1-D is performed locally on each block of the 2-D data. A ground object can occupy more than one row and/or more than one column of 2-D data. In addition, the properties in cross-range and range of an image are not the same. Thus, different conversion approaches may lead to different performance of an 1-D ANC and hence different change detection results. Among the considered approaches, the local concatenating approach is shown to provide slightly better performance in terms of probability of detection and false alarm rate. IEEE
The paper introduces an incoherent change detection algorithm with constant false alarm rate (CFAR). The algorithm is based on a CFAR detector that preceded with an adaptive noise smoothing filter while the input of the filter is preliminary changes retrieved from a subtraction of surveillance SAR image to reference SAR image. The algorithm is tested with 24 data sets provided by CARABAS. The average probability of detection calculated for 1200 deployed vehicles is up to 96% while the false alarm rate calculated for an area of 288 square kilometers is only 0.15 per square kilometer. In our test, the algorithm did not require a longer processing time than other wavelength-resolution SAR change detection algorithms. Based on these evaluations, the introduced algorithm is seen to be effective. © 2017 IEEE.
A statistical hypothesis test for wavelength-resolution SAR change detection can be derived with the bivariate distributions such as Rayleigh, Gamma and K. The paper investigates the stability of change detection results obtained with a statistical hypothesis test using bivariate Rayleigh distribution. Some practical issues concerning the implementation of the statistical hypothesis test such as scale parameter estimation, target magnitude assumptions and Bessel function calculation are also addressed. The statistical hypothesis test using bi-variate Rayleigh distribution are experimented with the data set containing 24 CARABAS II images. It is shown that beside the simplicity and efficiency, a statistical hypothesis test using bivariate Rayleigh distribution can provide very stable change detection results. © 2019 IEEE.
This paper introduces a method to reduce false alarms in wavelength-resolution synthetic aperture radar (SAR) change detection and aims at very high frequency-band systems like the Coherent All Radio Band System (CARABAS). The false alarms are usually caused by the elongated structures, such as power lines and fences, which stand out from the background. The responses of elongated structures are sensitive to flight path. The introduced method aims at minimizing the false alarms caused by the elongated structures and is based on the well-known adaptive processing mechanism, i.e., the so-called adaptive noise canceler (ANC) where a separate reference signal is required. The changes between measurements are considered by the input signal of ANC while the separate reference signal comes from the measurements without change. Hence, the method requires three SAR images associated with three measurements, with no changes between two of them. The reference data for the study are provided by CARABAS. The experimental results indicate that the method can reduce false alarms significantly and provide high probability of detection (≥98%). The experimental results also show that the method still works well even in the case where the flight tracks of the SAR system in the change detection measurements are slightly different.