AMTA Paper Archive

Planar near-field measurements are the usual choice when testing phased array antennas. NSI recently delivered a large state-of-the-art near­ field measurement system for testing a multi­ beam, solid state phased-array antenna. The critical sidelobe and beam pointing accuracy specifications for the antenna required that special attention be paid to near-field system design. The RF path to the moving probe was implemented using a multiple rotary joint system to minimize phase errors. Additional techniques used to minimize system errors were an optical probe position correction system and a Motion Tracking Interferometer (MTI) for thermal drift correction.

The ESAT-TELEMIC division at Katholieke Universiteit Leuven (KUL) has three antenna ranges: an indoor Far-Field range, an indoor planar Near-Field range and an outdoor Far-Field range. The positioning equipment is of a variety of manufacturers. The division launched an effort to modernize the range complex and add automatic measurement capabilities, while still retaining control of all three ranges from one control console and using one positioner controller, one angle readout and a single receiver to save costs. The system upgrade included some electrical refurbish­ ment of the positioning equipment and the replacement of all the old control and data recording equipment with Orbit Positioner Controller/Programmer, Power Control Unit and combined Near-Field and Far-Field software. Control of all three sites is achieved using a special Orbit Junction Box. With the new configuration all three ranges can be operated in fully automatic mode, one range at a time. The software package controls both Near-Field and Far­ Field measurements using compatible data formats and human interfaces.

For the past six years, ERIM has been studying RCS measurement error sources and processing methods by which these errors can be reduced. Typical errors that can be mitigated by processing techniques include near-field effects, multipath sources, and target support interactions. In this paper we briefly discuss image editing and spec­ tral decomposition methods which can be applied to error mitigation when the target sjze and bandwidth are suffi­ cient to resolve scattering centers. More details on these methods will be presented in other papers at this confer­ ence. We then describe in detail the netwcrk model approach which is best suited to applications where the target size is electrically small and the bandwidth is nar­ row. We show that the network model is a logical extension of the other techniques and discuss its application to error mitigation.

Coherent background subtraction is an established method of reducing additive range clutter in radar cross­ section measurements. In some measurement situations, it is neither practical nor convenient to directly make a coherent measurement of the range background. The Environmental Research Institute of Michigan has devel­ oped two methods of synthesizing background measure­ ments for the coherent subtraction of additive clutter in these cases. The first method synthesizes a background for measurements of pylon-supported targets by remov­ ing unterminated pylon returns using software gating. The second method improves background subtraction by compensating for phase drift between target and back­ ground measurements. In this paper, these methods of improving the performance and utility of background subtraction will be described and demonstrated on mea­ sured data.

Traditional range/doppler ISAR techniques have inherent geometric limitations. By using concepts of microwave holography and tomography, a vector-based k space approach allows a more generalized geometry of the sampled Fourier space. By constructing a complete annulus in the polar sampling space, arbitrary apertures up to 360 degrees can be processed for "full body" two dimensional images. This processing also typically exhibits better resolution. The algorithm relies on linear interpolation for potar­ Cartesian conversion. The general geometric formulation is also readily adaptable to arbitrary antenna configurations.

We discuss how RCS target depolariza­ tion enhances cross-polarization contamination, and we present a graphical study of measurement error due to depolarization by an inclined dihedral reflector. Error correction requires complete polarimetric RCS measure­ ments. We present a simple polarimetric calibration scheme that is applicable to reciprocal antenna radars. This method uses a dihedral calibration target mounted on a rotator. Because the calibration standard can be ro­ tated, there is no need to mount and align multiple sepa­ rate standards, and clutter and noise may be rejected by averaging over rotation angle.

Historically, radar imaging sensors have been divided into two categories, SAR and ISAR systems. Even though they are solving the same imaging prob­ lems the data collection environment is dramatically dif­ ferent between the two. Consequently, the particular waveforms selected for the two have been different. The primary waveform for ISAR RCS measurement systems is stepped frequency, while the FM-chirp (linear-FM) waveform has been used much more often in SAR applications. However, recently this boundary has been blurred, in that stepped frequency radars are being applied to long range dynamic measurements, long the domain of chirped waveforms, and conversely the chirped waveform has been applied to target RCS mea­ surements of both static and dynamic targets. This paper will address the system parameter tradeoffs involved in selecting between the two waveforms for two different applications; (i) near range static target imaging, and (ii) far range dynamic target imaging. The system parameter tradeoffs involve RF bandwidth, PRF, scene size, trans­ mitter power, doppler frequency spread of target, etc. The advantages, disadvantages, and inherent limitations of each waveform will be analyzed to yield a better understanding of the tradeoffs involved, and the data collection examples will further illustrate these tradeoffs for the two specific applications.

With modern range hardware, it is possible to per­ form ultra wide band frequency measurements with­ out changing the range configuration. This has not been possible with existing chamber antennas be­ cause they have been limited in bandwidth in or­ der to provide the desired illumination. In addition, these antennas have not considered scattering issues, even though one goes to great lengths to minimize reflections within a chamber. The rolled edge Slot­ line Bowtie Hybrid (SBH) antenna has been used for ultra wide band applications for many years. How­ ever, it can not meet the range scattering require­ ments due to its structure (large rolled edges). In this paper, a new R-Card version of the SBH an­ tenna is presented. It is fabricated by integrating resistive sheets (R-Cards) into the blended rolled edge concept so that both the ultra wide band and low RCS antenna features can be obtained simul­ taneously. Further, by employing resistive sheets, the R-Card SBH antenna can provide the desired constant beamwidth to fully illuminate the target zone. Measured and calculated results are presented to demonstrate the performance of this new antenna.

The identification of targets with radar is frequently based on a priori knowledge of the RCS characteristics of the target as a function of frequency and viewing angle. Due to the complex­ ity of most targets, it is difficult to predict their RCS signature accurately. Furthermore, complex and large reference libraries will be required for identification purposes. In most cases, a complete knowledge of the RCS is not required for successful identification. Instead, a target representation composed of the contributions of the main scattering centers of the target can be sufficient. This means that a corresponding target representation based on an estimation with Geometrical Optics (GO) or Physi­ cal Optics (PO) techniques will contain enough information for target identification purposes. In this paper, a new technique is described which is based on a reconstruction of the scattering centers. These are found at locations where the normal to the surface points in the direction of the angle of incidence. The RCS at these positions depends mainly on the local radii of curvature of the surface. Further­ more, PO and GO approximations are known as high-frequency techniques, assuming structures that are large compared to the wavelength. At low frequencies, which may be of interest for certain class of identification procedures, and for small physical radii of curvature, the RCS prediction is often difficult to determine numerically. Results from measurements show that this approach is also valid at lower frequencies for the classes of targets as mentioned, even for structures that are significantly smaller than the wavelength. As a consequence, it is expected that even complex targets can be represented adequately by the simplified model.

The amplitude of a point target observed in an ISAR image is equal to their free space RCS when effective sidelobe windowing is used. Likewise, its location in the image is identical to its actual location. The interpretation of observed amplitude and dimension of area targets is not as easy. The ISAR image of a rectangular flat plate formed by rotating it around its longer axis is significantly different from an ISAR image of the same plate rotated about its shorter axis. Both the amplitude and the size of the plate's image are different. In this paper, the theory of physical optics is reviewed in conjunction with the principles of ISAR processing to explain these differences.

Compact ranges have found wide application for antenna measurements, RCS measurements, and, most recently, for spacecraft payload measurements. Each of these ap­ plications requires certain special features of the range optics, positioning systems, electronics, and software. The system design of a compact range measurement sys­ tem for making all these types of measurements presents a number of challenges. This paper will discuss the system aspects of the design of a multi-purpose compact range facility. Items of inter­ est include the RF electronics design, the positioning sys­ tem design, the optimization of the reflector and feeds and the specialized software design.

ERIM is currently investigating several near-field to far-field transfonnations (NFFFfs) for predicting the far-field RCS of targets from monostatic near-field measurements. Each of the techniques uses approximate­ tions and/or supporting information to overcome the need for the bistatic near-field data which is required to rigorously transfonn a target's scattered field from the near zone to the far zone. Our focus has been on spheri­ cal near-field scanning, since this type of collection geometry is most compatible with existing RCS ranges. One particular NFFFT is based on the reflectivity approximation commonly used in ISAR imaging to model the target scattering. This image-based NFFFT is the most computationally efficient technique under con­ sideration, because, despite its theoretical underpinnings, it does not explicitly require image fonnation as part of its implementation. This paper presents an efficient discrete implementation of the image-based NFFFT, along with numerically-simulated examples of its perfonnance. The advantages and limitations of the technique will be discussed. A simplified version which applies to high aspect ratio (length-to-height) targets and requires only a single great circle (waterline) data in the near field is also summarized.

The RCS of extended objects measured in the near field is subject to errors induced by the spherical nature of the incident and scattered wavefields. A number of techniques have been applied to estimate far-field responses from results of monostatic near-field measurements. While the results indicate successful transformations for linear scatterers, the lack of a sound theoretical basis brings into question the appli­ cability to general objects. The paper explores the theoretical basis of the far-field transformation of RCS data and the consequence of the limited data obtained from monostatic measure­ments. The limitations of approaches reported to date [1-4] are explored from conceptual and physical con­ siderations with the goal of establishing reasonable expectations for practical methods. Examples using simulated and measured near-field data are presented to illustrate successes and failures of the algorithms in transforming results to far-field RCS.

Calibration of monostatic radar cross section (RCS) measurements is a well-defined process that has been optimized through many years of theoretical investigation and experimental trial and error. On the other hand, calibration of bistatic RCS measurements is potentially a very difficult problem; the range of bistatic angles over which calibration must be achieved is essentially unlimited and devising a calibration target that will provide a calculable scattering solution over the required range of bistatic angles is difficult, particularly for cross-polarized measurements. GTRI has developed a solution for amplitude calibration of both co-polarized and cross-polarized bistatic RCS, as well as a bistatic phase-calibration procedure for coherent systems.

Due to inherent cost, safety and logistical advan­ tages over dynamic measurements, Ground-to-Ground (G2G, aircraft and radar on tarmac) diagnostic radar measurements may be the preferred method of assessing aircraft RCS for signature maintenance. However, some challenging complications can occur when interpreting SAR imagery from these systems. For example, the effect of ground induced multi-path often results in the measurement of a significantly different image based RCS than would have been obtained by a comparable Ground-to-Air (G2A) or Air-to-Air (A2A) system. Although conventional 2-D SAR images are useful in determining the physical source (down-range/cross­ range) of scatterers, it is difficult at best to deduce whether an image pixel is a result of direct (desired) or ground induced multi-path (undesired) scattering. ERIM and MRC recently completed an experiment testing the utility of collecting and processing interfero­ metric (2-antenna) SAR radar data. This effort produced not only high resolution SAR imagery, but also a com­ panion data set, derived from interferometric phase, which helps to isolate the source (direct or multi-path) of all scattering within the SAR image. Additionally, the data set gives a measure of the physical height of direct scatterers on the target. This paper outlines the experiment performed on a RCS enhanced F-4 aircraft using a van mounted radar. Conventional high resolution imagery (down-range/ cross-range/intensity) will be shown along with down­ range/height/intensity and cross-range/height/intensity images. The paper will also describe the processing pro­ cedure and present analysis on the interferometric results. The unique motion compensation processing technique combining prominent point and motion mea­ surement instrumentation data, eliminates the need for a tightly controlled collection path (e.g. bulky rail sys­ tems). This allows data to be collected with the van driven somewhat arbitrarily around the target with side mounted antennas taking measurements at desired aspects.

For the past six years, ERIM has been studying RCS measurement error sources and processing methods by which these errors can be reduced. Typical errors that can be mitigated by processing techniques include near-field effects, multipath sources, and target support interactions. In this paper we briefly discuss image editing and spec­ tral decomposition methods which can be applied to error mitigation when the target sjze and bandwidth are suffi­ cient to resolve scattering centers. More details on these methods will be presented in other papers at this confer­ ence. We then describe in detail the netwcrk model approach which is best suited to applications where the target size is electrically small and the bandwidth is nar­ row. We show that the network model is a logical extension of the other techniques and discuss its application to error mitigation.

Image editing, a post measurement data processing technique, is an established method for the identification and reduction of non-target measurement artifacts like the target support system. The Environmental Research Institute of Michigan has applied this technique to data collected at the OSU "BIG EAR" VHF-UHF wideband compact range in order to remove or reduce target sup­ port interference and to extract selected target feature contributions to the RCS. In this paper, the application of the method to some BIG EAR measurements data is described and examples are shown which demonstrate the improvement in data quality and usability afforded by support contamination reduction and feature extraction techniques.

ERIM is investigating the use of modem spectral esti­ mation techniques for extracting (editing) desired or undesired contributions to RCS and ISAR measurements in two ways. The first approach involves using parametric spectral estimators to perform frequency sweep range compression and signal history editing, while the second involves using the associated stabilized linear prediction filters to extrapolate sweep data and perform "enhanced resolution" Fourier image editing. This paper summarizes our editing algorithms and illustrates RCS editing results using measurements of a conesphere target contaminated by a metal rod and foam support. The reconstructed "clean" conesphere measurements are compared quantitatively against numerically simulated ground truth. Editing was performed using three bandwidths at two center fre­ quencies to provide insight into the impacts of nominal resolution and scatterer amplitude variation with fre­ quency on editing efficacy, and to assess the degree to which superresolution algorithms can offset reduced nominal resolution.

Spatially Variant Apodization (SVA) [l] is nonlinear image domain algorithm which effectively eliminates finite-aperture sidelobes from SAR/ISAR imagery without degrading mainlobe resolution, unlike traditional methods of sidelobe suppression (e.g. Taylor weighting). Dezellum et. al. [2] demonstrated at the 16th AMTA symposium the benefits of SVA for improving RCS analysis of ISAR data. The purpose of this paper is to show that robust super-resolution via bandwidth extrapolation can be obtained in a relatively simple, straightforward manner using SVA, providing further improvement in RCS measurements from SAR/ISAR data. This new super-resolution algorithm (called Super-SVA) can extrapolate the signal bandwidth for an arbitrary set of scatterers by a factor of two or more, with a commensurate improvement in resolution. Super-resolution techniques have been traditionally limited to problems where a-priori knowledge is available and/or the scene content is suitably constrained. Using Super-SVA, no a-priori knowledge of scene content is required. Super-SVA exploits the fact that SVA applied to an image results in finite image-domain support on the scale of the system resolution for an arbitrary set of complex scatterers. Extrapolation of the frequency-domain signal data is then simply a matter of applying frequency-domain inverse amplitude weighting. The fidelity of the deconvolution process can be improved by embedding the original signal data in the extrapolated data and performing further iterations of the process.

The U.S. Air Force 46 Test Group, Radar Target Scattering Division (RATSCAT), at Holloman AFB, NM, in conjunction with the US Army, Navy and Georgia Tech Research Institute (GTRI), has developed a concept for a bistatic coherent radar measurement system (BICOMS). It will be used to measure both the monostatic and bistatic RCS of targets, as well as create two-dimensional images of monostatic and bistatic signature data. It will consist of two mobile radar units, each of which is capable of simultaineously collecting coherent monostatic and bistatic RCS data. This paper will cover the systetn design specificatiovs, layout and design of equipment, and discuss the operating parameters for the radar (power, antenna sizes, sensitivities, timing, etc.).