however worsens again for PRFs larger than 2.25 kHz. The reason for this behaviour is the spatial sample distribution . For the given geometry a PRF of about 2.2 kHz leads to a uniform sampling of the received signals. Increasing the PRF further increases the non-uniformity of the sampling which compensates the benefit of a larger PRF and even further deteriorates the AASR performance. The PRF of the actual acquisition was 2 kHz leading to a sampling uniformity of 0.6 for a scale between 0.0 and 2.0 (uniform sampling for 1.0). This results in a theoretical AASR performance of about -15 dB for the reconstructed signal. The single channel achieves about -5 dB AASR. In order to compare the simulated AASR values with the results obtained from Figure 6 the simulated values have to be adjusted by -3 dB. This is necessary as the simulationresults consider ambiguities in both azimuth directions whereas the azimuth profiles are dominated by a single sided ambiguity, i.e., the strong backscatter region of the land side. The simulation levels for a comparison are thus -8 dB for the single channel and -18 dB for the reconstructed data. The single channel AASR value agrees very well with the actual value derived from the azimuth profile. However, for the reconstructed signal the simulation predicts an AASR of -18 dB but the image shows -14 dB. This deviation might by caused by effects not jet considered within the simulation framework or by inaccuracies during the multi-channel reconstruction of the experimentaldata. Additionally, the residual ambiguity in Figure 6 is very close to the noise floor which complicates an assessment of lower ambiguity levels.
Fig. 4 shows the results for a reconstruction from non- uniform DPCA sampling. For better illustration, an extended scatterer has been chosen in this simulation. Furthermore, independent white noise has been added to each sub-aperture signal to account for differences in the two receiver channels. It becomes clear that the signal from each channel is strongly ambiguous (left) while the coherent reconstruction of the two sub- sampled SAR signals will provide a efficient ambigu- ity suppression (right). The algorithm has also been tested with real SARdata acquired by the DLR E- SAR system. Since there is currently no displaced phase centre mode in the E-SAR system, monostatic data were used in this example. The data, which have an original sampling frequency f s , were filtered with
Due to the all-weather data acquisition capabilities, high resolution space borne Synthetic Aperture Radar (SAR) plays an important role in remote sensing applications like change detection. However, because of the complex geometric mapping of buildings in urban areas, SAR images are often hard to interpret. SARsimulation techniques ease the visual interpretation ofSAR images, while fully automatic interpretation is still a challenge. This paper presents a method for supporting the interpretation of high resolution SAR images with simulated radar images using a LiDAR digital surface model (DSM). Line features are extracted from the simulated and real SAR images and used for matching. A single building model is generated from the DSM and used for building recognition in the SAR image. An application for the concept is presented for the city centre of Munich where the comparisonof the simulation to the TerraSAR-X data shows a good similarity. Based on the result ofsimulationand matching, special features (e.g. like double bounce lines, shadow areas etc.) can be automatically indicated in SAR image.
airborne bistatic synthetic aperture radar (SAR) experiment, conducted early November 2007, using the German satellite TerraSAR-X as transmitter and the German Aerospace Cen- ter’s (DLR) new airborne radar system F-SAR as receiver. The importance of the experiment resides in both its pioneering char- acter and its potential to serve as a test bed for the validation of nonstationary bistatic acquisitions, novel calibration and synchro- nization algorithms, and advanced imaging techniques. Due to the independent operation of the transmitter and receiver, an accu- rate synchronization procedure was needed during processing to make high-resolution imaging feasible. Precise phase-preserving bistatic focusing can only be achieved if time and phase syn- chronization exist. The synchronization approach, based on the evaluation of the range histories of several reference targets, was verified through a separate analysis of the range and Doppler contributions. After successful synchronization, nonstationary fo- cusing was performed using a bistatic backprojection algorithm. During the campaign, stand-alone TerraSAR-X monostatic as well as interoperated TerraSAR-X/F-SAR bistatic data sets were recorded. As expected, the bistatic image shows a space-variant behavior in spatial resolution and in signal-to-noise ratio. Due to the selected configuration, the bistatic image outperforms its monostatic counterpart in almost the complete imaged scene. A detailed comparison between monostatic and bistatic images is given, illustrating the complementarity of both measurements in terms of backscatter and Doppler information. The results are of fundamental importance for the development of future nonsyn- chronized bistatic SAR systems.
level and the other based on image data level. The approaches were tested and validated with simulated as well as with real TSX data. Both the simulated scalloping and the measured scalloping are very small (0.2 dB). Therefore, despite the very good absolute antenna model accuracy ( ≤ 0.2 dB), the relative accuracy of the azimuth patterns is the limiting factor for the scalloping correction. However, generally, scalloping correction is desired, since larger maximum steering angles imply also a larger scalloping. Both correction methods show a very good performance with TSX dataand yield comparable results. The raw-data-based method is more promising, as it can be applied directly to the data prior to SAR processing, whereas the image-data-based method requires an additional simulation step and a SAR postprocessing correction. Additionally, the raw data correction can be easily verified by an equalized raw data profile, which provides the opportunity for a data- based calibration approach. In order to correct for a very small systematic bias in the gain of the backward-steered azimuth beams within the absolute TSX antenna model accuracy, several averaged scalloping-corrected raw data profiles were used to derive a correction function and to radiometrically correct the TOPS acquisition. This avoids a recalibration of each single beam and can be also used for an efficient calibration of all steered azimuth beams of a SAR system.
-700 -600 - - - -6.72 0.01 -6.73
3.2 X-band data
The X-band data were collected using two identical monostatic CW radar systems which were deployed at Node 2 and Node 3, as shown in Fig. 1. Fig. 11 presents examples of spectrograms for both VV polarised and HH polarised data which were collected simultaneously to the S-band data shown in Fig. 3 and Fig. 5. These were calculated using the same window overlap as for the S-band data (95%) and a quarter of the window length (0.15 s) in order to allow a comparison, with the same Doppler shift to frequency resolution ratio, between the S-band and X-band data. Results show that the micro-Doppler signature of the turbine is spread across a wider range of Doppler frequencies which is, as expected, approximately four times broader than that of the S-band data (being the X-band carrier frequency roughly four times higher). The signature at node 2 shows a lower Doppler spread in both VV and HH polarisations. Although each X-band node worked as independent monostatic node, these results seem to show that multiple nodes surveying an area from different spatial locations can be affected differently by wind farm clutter depending on polarisations and aspect angle between node and yaw axis of the turbine. In the previous section similar effects were observed for S-band monostatic vs bistatic data, i.e. the possibility that one radar node is affected by more favourable clutter when multi-perspective views from multiple nodes are used. The contribution of the blade components moving away from the radar generate a stronger return in HH polarisation as for the S- band data. For both polarisations the blade flashes reach higher negative Doppler shifts.
baseline variation over the data take is small. However, the time variant along-track baseline and its impact on the reconstruction is still subject of on-going investigations. In the lower left part of the images a floating ice sheet and its ambiguity are visible. This particular ambiguity experiences almost no suppression in the reconstructed image compared to the single channel image. Addition- ally, the intensity of the ice sheet itself is modulated in az- imuth direction in the right scene. This effects are caused by a motion of the ice sheet between both acquisitions, which are separated by about ten seconds. The motion introduces an additional phase term which is not com- pensated during the processing and therefore affects the reconstruction performance.
The paper demonstrates the applicability of the multi-channel reconstruction algorithm to measured multi- channel X-band data. Up to four individual receiving channels, each sub-sampled, were combined to a single channel, ideally free of aliasing but in reality showing residual azimuth ambiguities caused by channel imbal- ances. Consequently, different channel balancing methods were presented, demonstrating improved ambi- guity suppression. In a first step, channel balancing was applied to the data before decimation, in order to show the impact of channel imbalances and to allow for a comparisonof different approaches. As these Doppler domain methods fail when applied to sub-sampled data after decimation, a second step derived a new channel balancing approach applicable to aliased data. This histogram-based method in range-Doppler allows for compensating a constant phase offset which represents the main cause of imbalances, by this enabling a clearly better balancing and improved reconstruction results. The approach is comparable to a time domain synchronization method , after a squint-angle pointing adaptation to account for the Doppler centroid. Based on these promising results, future work will further elaborate channel balancing techniques for sub-sampled data in order to derive suitable calibration strategies for future multi-channel systems. References
This paper has examined novel SAR imagery phenom- ena arising from vibrating targets and multi-path effects that result in degraded imagery. A mathematical foun- dation for their origin has been presented and numeri- cal experiments of realistic scenarios simulated. Lab- oratory experiments replicating, as far as possible, the practical scenarios and the resultsof the data gathered were in good agreements with calculated. This work validates our hypothesis and the software tools and lab- oratory techniques will allow the investigation of more through-wall SAR phenomena and artefacts.
The generation of high-resolution digital elevation models (DEM) requires the knowledge of relative phases within a few degrees to avoid low-frequency modulation of the DEM in azimuth . In the TanDEM-X mission, synchro- nization with this level of accuracy was achieved by ex- changing radar pulses between the satellites through a di- rect microwave link . The direct link involved dedicated transmit and receive hardware and a total of six dedicated antennas covering the full solid angle .
For the experimental analysis of the dependence of the beam propagation on weather conditions, the DLR oper- ates a free transmission laser test range at Lampoldshausen, Germany. It consists of two stations, the transmitting (TS) and the receiving (RS) one confining a 135-m-long pathway with a beam path 1 m above asphalt ground. Optical turbu- lence has been measured by a surface layer scintillometer (SLS 20-A, Scintec AG, Germany). A more comprehensive description of the optical test range including installed sen- sors continuously monitoring the local atmosphere can be found in Ref. [ 17 ]. The simultaneous characterization of the laser beam is performed with sensors inside of the TS and RS. These measurements address power, intensity distribu- tion, and jitter of the beam.
The meteorological and geological data presented in this work has been acquired at two stations (see Fig. 1a) close to Zagora (30°19´50´´ N, 5°50´17´´ W) and Missour (32°53´46´´ N, 4°06´37´´ W), both in Morocco. The two stations belong to the enerMENA meteorological network [17, 18]. The reflector samples are mounted at a tilt angle of 45° facing south in a height between 1 and 1.5 m above ground level. The direct normal irradiance (DNI) is measured with a rotating shadow band irradiometer (type: RSP-4G manufactured by Reichert GmbH) in Zagora. In Missour, a solar tracker with mounted CHP1 pyrheliometer by Kipp & Zonen (Delft, Netherlands) is installed. Temperature and relative humidity are measured by a Campbell Scientific CS100 (Logan, Utah, USA) and the wind velocity and direction at 10 m height is measured with a #40C and a #200P from NRG (Hinesburg, Vermont, USA), respectively. An EDM164 manufactured by GRIMM Aerosol Technik GmbH&Co.KG (Ainring/Germany) (Fig. 1c) has been mounted at a sampling height of around 1 m. This instrument counts airborne particles with diameters ranging from 0.25 to 1200 µm by light scattering optics. It differentiates between 31 size channels where the largest one counts all the particles with a diameter larger than 32 µm. The evaluation will only be performed up to particle diameters of 31 µm, in order not to lose accuracy because of the poor sampling efficiency of this instrument at larger particle sizes that other scholars pointed out already .
When ordering a scene from SpotLight mode, a user is guaranteed that the delivered product cov- ers at least 10 km x 10 km in SL and 5 km x 5 km in HS mode around the ordered scene centre coordinates. As for all SARdata acquisitions, the Doppler centroid and its variations strongly influ- ence the raw data azimuth start and stop times. Whereas the duration ofdata taking in StripMap or ScanSAR configuration can be quite easily extended by considerable margins, the situation is by far more complex for the SpotLight mode due to the azimuth beam steering. The automatic TMSP check of the deviations between the ordered and the processed scene confirms that the SL and HS products are within their specified extent and location accuracy. Note that the 300 MHz HS products show a reduced range extent (still centred around the ordered scene centre) due to the on-board echo buffer limitations, in some cases also a shortened azimuth length.
Comparison Between Simulation- and Test Resultsof an Observer-Controlled MAGLEV Vehicle on Elastic Guideway
8 Schriftenreihe der Georg-Simon-Ohm-Hochschule Nürnberg
Taking the guideway damping with c = 10 N s m into account and assuming guideway natural frequency ⋅ / changes in the scope of 1Hz~100Hz, the characteristic roots of the matrix A r are calculated. The maximum real parts of the roots for the different frequencies are shown in the Fig.6. The figure shows that with the increase of the natural frequency, the maximum real parts decrease and become negative when the frequency or the stiff- ness is above a certain value, which means the system can be stable if the guideway stiffness is hard enough. That is why the guideway beams are built as hard as possible in practice. The time-domain simulation result (for a natural frequency of the guideway of 35Hz) shown in the Fig.7 indicates that it takes about 0.2s for the system to reach the steady state and the maximum overshoot is about 8%.
different receiver channels have been used. Switching between both channels gives a sampling blank area of 1.6 µs and is performed with a 7693.25 Hz cadence; maximal real sampling rate using this configuration is 250 MHz. Selected transmitted chirp bandwidth is 100 MHz and down converting reference in receiver matches the nominal value of TSX carrier for this acquisition, 9.65 GHz. Chirp duration is 33.189 µs. To increase footprint overlapping time, TSX antenna is steered in azimuth to perform a spotlight illumination of the scene. F-SAR is receiving in regular stripmap mode. A high PRF of 5920.59 Hz is chosen to guarantee a high along-track oversampling rate of the bistatic data. Since the imaged scene is limited by F-SAR antenna pattern, no range ambiguities are expected to arise in bistatic image; range ambiguities are however expected in monostatic image, even worsened by the atypically shallow look angle used in the satellite antenna pattern. Transmitted peak power is 2.01 kW. Taking into account the difference in transmitted power, antennas and free space losses of this bistatic acquisition compared to a pure airborne monostatic one, F-SAR receiver gain has been increased 22 dB with respect to monostatic operation. Three X-band transponders used for TSX calibration during its comissioning phase are used as reference targets on ground. Their exact position is measured with GPS; these targets will be essential to synchronise the data in further processing steps.
3. COMBINED USAGE OF SEVERAL DSM The common method up to now is DEM generation from scratch, i.e. a completely new DSM is generated from a new satellite data set. On the other hand, worldwide DEM coverage is already available. Since 1996 GTOPO30 is accessible, a global DEM with a horizontal grid spacing of 30 arc seconds (approximately 1 km). Dataof the SRTM mission refined the globally available DEM (80% of Earth's land mass) to a resolution of 3 arc seconds and much better height accuracy. The observed scene is unique, so it seems natural to obtain only a single DEM instead of having several individual DEM. Thereby, the density of reliable information increases resulting in more precise DEM than with individual DEM. Here two different methods are presented to produce combined DEM. Input to the DEM combination methods are the best available independent derived DEM and/or point clouds. For the DEM from optical data, the SPOT-5 HRS DEM derived during the ISPRS/CNES assessment program was used. (Reinartz et al., 2006). The global shift between all available DSM and the reference DEM is estimated via iterative least squares adjustment and the DSM are moved accordingly in X, Y, and Z. 3.1 DEM fusion
reducing the length of the ’synthetic aperture’ . Building upon that, the goal of the research for this work was to present a low cost realistic SAR implementation using minimal hard- ware as a proof of concept. The signal processing algorithm has been developed so that it reflects the realistic real-time setup if it should be implemented in a car. The presented algorithm is faster and therefore well suited for a real-time implementation. Also the gaps between signals in azimuth (due to hardware restrictions, discussed in Section II-B) have been filled by reconstructing the missing signals with the help of compressed sensing . Additionally the processed images are verified by comparison against the same measurements processed through the backprojection algorithm .
Radar (SAR) data sets, recent spaceborne sensors reach their limits in terms of resolution, swath width and repeat cycle. In addition, fully polarimetric operation reduces the maximum swath width approximately by the factor of two. To solve SAR inherent limitations, a new generation of sensors with multiple transmit and multiple receive channels (MIMO) and digital beam-forming (DBF) capabilities is suggested. Using a quad- polarized MIMO-SAR, transmitting simultaneously orthogonal waveforms in horizontal and vertical polarizations, enables data acquisitions without any reduction of the swath width or resolu- tion. In this paper a concept is described, which uses an advanced filtering and processing method to separate the transmitted signals in the receiver and to measure all four parameters of the scattering matrix simultaneously. Ground-based MIMO- SAR measurement results are presented, which serve as a first verification of the suggested technique with an extended, non point-like target.