Welcome to the AMTA paper archive. Select a category, publication date or search by author.
(Note: Papers will always be listed by categories. To see ALL of the papers meeting your search criteria select the "AMTA Paper Archive" category after performing your search.)
In previous AMTA Symposia, the Air Force Research Laboratory reported on a successful effort to fabricate, measure, and predict the precise radar cross section (RCS) for various cylindrical calibration targets [1]. In this paper, we apply what we have learned about calibration cylinders to the study of a 3.048 meter ogive body of revolution. Recall that an ogive is simply the arc of a circle spun on its axis. The radar signature of this shape is extremely small in the direction of the "point", even at low frequencies. A few years ago, AFRL had the subject ogive built for an RCS inter-range comparison between AFRL and the NRTF bistatic RCS measurement system [2]. In this paper, we utilize this ogive body to assess both the quality and accuracy of VHF RCS measurements and predictions performed using multiple calculation schemes. In the end, reconciling the ogive measurements and predictions led us to reassess how composite objects are "conductively coated" to simulate a perfect electric conductor. This insight resulted in refinements in the process for measuring and predicting the ogive at low frequencies where electrical size and electromagnetic skin depth considerations are important.
In order to better estimate the uncertainties in measured RCS for the Boeing 9-77 Compact Range, we study the responses from three high-quality objects, i.e., two ultraspheres of 14” and 8” in dia., plus the 4.5" squat-cylinder, each supported by strings. When calibrated against each other in pairs, the differences between measured RCS and predicted values are taken as the uncertainties for either object. Two standard-deviations from the target, reference, and background, as computed from repetitive sweeps, are taken as the respective uncertainties for the signals. Using the root-sum-squares (RSS) method, the error bars are found to be between + 0.1 to 0.2 dB for most of the frequency F, from 2 to 17.5 GHz. We also analyze the responses from a thin steel wire (dia. 0.020"), supported by fine fishing strings (dia. 0.012"), at broadside to the radar. When the ‘wire and string’ assembly is oriented vertically, the HH echo from the 3-ft metal wire alone happens to be comparable to the HH from the 30-ft dielectric strings. Varying with F4, the combined RCS in HH for the assembly spans a wide range of 38 dB from 2 to 18 GHz. The error bounds are found to bracket the measured traces even when the signals are barely above the noise floor.
Future scientific and earth observation instruments as MASTER, PLANCK and HERSCHEL of ESA/ESTEC are working in the sub-millimeter wave range. For measurement of the instruments, a study named ADMIRALS was performed, mainly to identify the most suitable test facility, procure transmit and receive modules and perform measurements up to 500 GHz. The CCR 75/60 of Astrium GmbH, Ottobrunn, was selected for the facility calibration and the pattern verification with an Representative Test Object (RTO). The measurements were performed in three different frequency bands between 200 and 500 GHz. The mmwave transmit and receive modules were designed, manufactured and tested by Radiometer Physics GmbH (RPG). A cost efficient design was achieved by a modular concept. Within this paper, the design and realization of the modules as well as most characteristic performance parameter will be presented.
In this paper, the authors propose an improved Rotatingelement Electric-field Vector (REV) method taking into account amplitude and phase error of phase shifters in order to achieve more precise calibration. The conventional REV method has been used in order to determine and/or adjust amplitude and phase of electrical field radiated from each antenna element -element fieldin phased array antennas. However, amplitude and phase deviations due to phase shifter errors, and so on, reduce the measurement accuracy because the conventional REV method assumes no deviation. On the other hand, the proposed REV method can evaluate element fields without error and error electrical fields -error fields- due to phase shifter errors in each bit, by measuring both amplitude and phase value of array composite electrical field. In a simulation for a 31- element array with 5-bit phase shifter, the evaluated element fields and error fields agree well with the expected values. This result shows that the proposed method allows the phased arrays to be calibrated more accurately as considering phase shifter errors.
A thermal technique for the remote calibration of phased array radar antennas is proposed in this paper. The technique is based on infrared (IR) measurements of the heat patterns produced in a thin planar detector screen placed near the antenna. The magnitude of the field can be measured by capturing an isothermal image (IR thermogram) of the field with an IR imagining camera. The phase of the field can be measured by creating a thermal interference pattern (IR/microwave hologram) between the phased array antenna and a known reference source. This thermal imaging technique has the advantages of speed and portability over existing hard-wired probe methods and can be used in-the-field to remotely measure the magnitude and the phase of the field radiated by the antenna. This information can be used to calibrate the individual elements controlling the radiation pattern of the array.
The level of multiple reflections in near-field antenna measurements is an important issue in a measurement error budget. Traditionally, the interactions between the test antenna and the measuring probe have been reduced by covering the probe mounting structure with absorbing material. In this paper, a novel approach to alleviating the problem is discussed. This implies the use of a skirt to act as a shield against the mounting structure behind the probe, thereby eliminating the need for an absorber, which is a fragile material when exposed to wear and tear. This also has the added advantage that probe calibration data will not depend on a particular absorber that must be considered as an integral part of the probe. With a suitable design of the skirt, the level of multiple reflections can be reduced, whilst at the same time maintaining the pattern of the probe in the boresight direction unchanged. Prototypes of probes for 20 GHz and 30 GHz have been manufactured and tested, and excellent agreement between experimental results and theoretical predictions has been observed.
Compensated Compact Ranges (CCR) represent a high standard of state-of-the-art test facilities with a fast and real time measurement capability up to the submm wave range. Future scientific and earth observation instruments of ESA/ESTEC such as MASTER, PLANCK and HERSCHEL are working within this frequency ranges and require a high measurement accuracy for large antenna apertures. Within the ADMIRALS study for ESA/ESTEC, transmit and receive modules up to 500 GHz and an appropriate large offset reflector antenna with precise surface accuracy in form of a Representative Test Object (RTO) were applied. Related tests in the CCR 75/60 of Astrium were performed in order to qualify the test facility and verify the antenna measurements with theoretical pattern calculations. The present paper shows measurement results with the highly accurate Plane Wave Scanner (PWS) of Astrium GmbH and the RTO. Through the measurements performed, the accuracy of the plane wave field as well as pattern accuracy in the quiet zone of the CCR 75/60 have been qualified up to 500 GHz.
A method is presented for calculating the gain of corrugated conical horns. It is based on basic symmetry conditions of circular or conical waveguide mode fields. This formulation allows to derive the radiation pattern over a complete sphere form two principal polarization patterns (E- and H-plane patterns). This method can be applied for both theoretical or experimental patterns, respectively. The theory has been verified experimentally with measurements carried out on two different ranges. The results agreed within 0.05 dB or less in all situations.
Accurate gain measurements using any measurement technique require a calibrated gain standard, and the uncertainty in the gain of the standard is usually the largest term in the error analysis. To reduce the uncertainty, gain standards are often calibrated using a three- antenna measurement technique and the resulting gain values are generally certified to have an uncertainty of approximately 0.10 dB1-11. For near-field measurements, the gain standard may be the probe that is used to obtain the near-field data or it may be a Standard Gain Horn (SGH). Since the calibration of the gain standard is time consuming and often costly, it is desirable to verify that the gain of the standard is stable over long periods of time. This paper will describe tests to verify the gain stability of the standard and will also illustrate the terms in the error analysis that have the major effect on the uncertainty of any near-field gain measurement. With proper attention to the major error terms, the stability of the gain standard can be verified to approximate the original calibration uncertainty.
A fundamental part of the work of the BAE SYSTEMS Advanced Technology Centres Materials Group at Towcester (UK) is the microwave characterisation of the electromagnetic parameters of lossy materials. This paper describes a Quasi-optical microwave system for the free space measurement of material parameters in the frequency range 5 GHz to 18 GHz. The system employs two spherical reflectors which are illuminated from the side by gausian beam forming antennas. This produces a well defined parallel beam between the reflectors. The 5 GHz ro 18 GHz frequency range is covered in three bands with three pairs of corrugated feed antennas. An advantage of this system is that the beamwaist diameter (or illumination area) is essentially the same for each of the three frequency bands The measurements are taken using a vector network analyser under computer control. The parallel beam enables a “Through,Reflect,Line” calibration technique to be used. After calibration the sample under test is placed in the beam (mid way between the reflectors) and the four microwave ‘S’ parameters are recorded automatically in complex form. The permittivity, permeabilty or lumped admittance (if the sample is very thin <ë/50) for the material are then determined from the ‘S’ parameters. The operation and performance of the system is discussed and some material parameter measurement results are given.
Microwave holography is a popular method for diagnosis and alignment of phased array antennas. Holography, commonly known in the near-field measurement community as "back transformation", is a method that allows computation of the primary (aperture) fields from the secondary (far-zone) fields. This technique requires the far-zone fields to be known over a complete hemisphere and adequately sampled on a regular spaced grid in K-space. The holography technique, while known to be mathematically valid, is subject to errors just as all measurements are. Surprisingly, very little work has been done to quantify the accuracy of the procedure in the presence of known measurement errors. It is unreasonable to think that the amplitude and phase of the array elements can be trimmed to better than the uncertainty of the back-transformed amplitude and phase. This makes it difficult for an antenna engineer to determine the achievable resolution in the measurement and calibration of a phased array antenna. This study reports the results of an empirical characterization of known errors in the holography process. A numerical model of the near-field measurement and holography process has been developed and many test cases examined in an effort to isolate and characterize individual errors commonly found in planar microwave holography. From this work, an error budget can be developed for the measurement of a specific antenna.
Undesired antenna motion can significantly degrade SAR and ISAR image quality on an instrumentation radar operating in an outdoor or uncontrolled environment. Antenna vibration on the order of only a few hundredths of an inch at X-band frequencies can degrade performance to the point that one cannot reliably differentiate between the true and false peaks in the radar image. This paper describes a motion compensation technique that utilizes the measurements from an auxiliary antenna pointing at a stationary target. This "Phase Monitoring Subsystem" accurately records the linear antenna motion profile, which can then be used for compensation. Data collected at the US Naval Undersea Warfare Center (NUWC) Fisher Island Test Facility on a calibration target demonstrate that this compensation technique can reduce image artifacts by more than 20 dB.
Recently there has been a large effort to improve RCS range performance. Reducing errors associated with an RCS measurement requires the identification of stray signal sources, highly accurate calibration, and an understanding of the target mount interactions. This paper will illustrate the potential errors resulting from target mount interaction. A complex RCS target of generic shapes was designed to illustrate target support interactions. Target features include a front wedge shape, a rear circular shape and a vertical fin. All the target features are separable in time using a 2-18 Ghz measurement system. The target features were designed to strongly interact with the ogival pylon. Measurements using the metal ogival support show strong interactions resulting from the shadowing effect produced by the metal ogival pylon. The measurements were repeated using a foam column mount. Since the foam column interacts much less strongly than the metal ogive, the foam column results are much more accurate.
A diagnostic/prognostic method for phased-array antennas has been developed which uses a single, fixed position RF probe to detect and identify faulty array elements as the array operates normally. After system calibration, zero array down time is required. A fiberoptic RF probe which allows implementation of the technique while negligibly affecting array operation and performance, has also been developed. The system has been demonstrated both in various computer simulations of arrays to over 1000 elements and in recent experimental tests at NSWC, Crane, IN. Identified faults include array elements which were off, stuck at constant phase, low in power, including both single faults and large numbers of simultaneous faults. An RF radiating probe (a fiber-optic version of which has also been demonstrated) can be used to diagnose array receive mode operation. Results of both the simulations and the tests are reported along with the design of the fiber-optic probe.
A model-based approach is presented to estimate the desired planar wavefront (DPW) component in range probe data. The estimated DPW component at the center of the quiet zone can be used effectively to calibrate frequency domain range probe data. The calibration is required when the range probe data is used for stray signal analysis. Using a simulated range probe data set and an experimental range probe data set, it is shown that the model-based DPW estimate is better than the DPW estimate obtained using simple smoothing. This is especially true at low frequencies where the quiet zone of a range is limited to 5-6 wavelengths.
For greatest efficiency and accuracy in measuring patterns of a circularly polarized antenna on a planar near field range (NFR), a recommended procedure is to use a fast switched, dual circularly polarized probe. With such equipment one obtains complete pattern and polarization data from a single scan of the antenna aperture. For our task of measuring high gain shaped beam apertures, measurement efficiency is further improved by using a moderately high gain (about 12 dBi) probe that has been accurately calibrated for patterns, polarization, and gain over the test frequency band. Such a probe allows scan data point spacing to be typically at least one wavelength, thus keeping scan time minimized with acceptably small aliasing (data spacing) error. The measured near field amplitude and phase data is transformed via computer to produce the angular spectrum that is further processed to remove the effect of the probe patterns, i.e. probe correction. The final output is a set of (principal and cross) circular polarized far field patterns. However on one occasion, due to fast breaking changes in requirements, we were unable to obtain a calibrated circular polarized probe in the available time. For this test we used an available calibrated 12 dBi fast-switched dual linear-polarized probe with software capable of processing principal and cross circular-polarized far field patterns. As anticipated, we found from preliminary tests that the predicted low cross-polarized shaped beam pattern was not achieved when using the calibrated fast Ku band probe switch. Further tests showed the problem to be due to small errors in calibration of the probe switch. This paper will discuss test and analysis details of this problem and methods of solution.
A radar transceiver operating at 1.56 THz has recently been developed to obtain coherent, fully polarimetric W-band (98 GHz) RCS images of 1:16 scale model targets. The associated optical system operates by a scanning a small focused beam of swept frequency radiation across a scale model to resolve individual scattering centers and obtain the scaled RCS values for the centers. Output from a tunable microwave source (10 - 17 GHz) is mixed with narrow band submillimeter-wave radiation in a Schottky diode mixer to produce the chirped transmit signal. Two high-frequency Schottky diode mixers are used for reception of the V-pol and H-pol receive states, with a fourth mixer providing a system phase reference. The full 2x2 polarization scattering matrix (PSM) for each resolved center is obtained following off-line data processing. Measurement examples of five simple calibration objects and a tank are presented.
During the last 6 years scientists at NIST have been focusing on radar cross section (RCS) measurements to improve RCS uncertainty analysis, and to develop new measurement and calibration artifacts and procedures. In addition, NIST has been asked to provide technical support to the DoD RCS self-certification effort. In this talk I review the technical accomplishments of the program, and will make suggestions for future research to improve RCS calibration and measurement technology. I will also present the structure of the certi fication process, and discuss NIST's role in the ongoing certification activities.
As a result of Government and Industry RCS Teaming, initial RCS range certification exercises are underway. One critical element of certification exercises is the modeling and characterization of error terms according to the unique properties and requirements of individual RCS ranges, and the development of a method for propagating these errors into overall RCS measurement uncertainty. Previously, we presented the statistical model for the case where errors are grouped into multiplicative and additive classes, as well as a robust methodology for the propagation of errors in both the signal space and RCS (dBsm) domains [1-3]. Initial data at the National RCS Test Facility (NRTF) RAMS site located in the White Sands Missile Range near Holloman AFB, NM, have been collected for range certification exercises. Preliminary analysis has been accomplished on certain dominant error terms for calibration uncertainty characterization only. A general approach [7] has been followed here, with the exception that multiplicative and additive error terms are treated separately. In addition, only variance effects are treated (not bias). This paper is a status of work in progress. The ultimate goal of this work is the full implementation of previously described concepts [1-3]. We plan to demonstrate an improved ability to capture the effects of both error bias and variance (as has been demonstrated mathematically to date) using a more complete set of data collections.
This is a report on work in progress to understand the wide variations in sphere calibration data observed on dynamic radar cross section measurement ranges. The magnitude of these fluctuations indicate an uncertainty of greater than 2 dB in some cases. The range of fluctuations in the received power (which is well beyond fluctuation due to received noise) underlines the need for a thorough understanding of sources of uncertainties in dynamic radar cross section measurements. In addition to the fluctuations, we observe a systematic error with respect to the mean of the data segments, possibly due to drift, pointing errors and I or target-background interactions. Understanding the error mechanisms in these measurements allows us to reduce the overall uncertainty and to improve data quality.