This paper systematically evaluates the performance potential of ultra-wideband (UWB) radar systems for Moving Target Detection (MTD) in strong clutter. Based on a general data model for UWB multiple phase waveforms, we first derive the waveform-based optimum UWB processor which represents the maximum achievable performance of UWB systems. The UWB gains/losses are individually identified by careful performance comparisons with narrowband systems. Our study results show that UWB systems potentially have a significant MTD performance gain, in addition to the well-known clutter and fluctuation reduction. This additional gain indicates that the subclutter visibility performance of UWB surveillance systems is severely underestimate in the existing literature.

It is well documented that multipath interference at low grazing angles is an issue for detection and tracking. As the conventional systems, a propagation factor is derived which quantifies system loss due to amplitude fading and phase distortion caused by the interference of 'free space' signals with multipath. For an ultrawide bandwidth (UWB) monocycle waveform operating at UHF, the earth's surface generally appears smooth and supports dominant coherent (specular) reflection. For interference patterns dominated by specular multipath, a closed form of the gain/loss (G) expected at the output of a matched filter (used as the UWB radar's receiver model) was derived and evaluated for a frequency dispersive and non- dispersive terrain model. The frequency dependence of the gain/loss function and the relative weighting of the multipath fields for the scenario analyzed was found to be dominated by the line spectrum of the UWB waveform S(f). In addition, for multipath to have any significant effect in a single monocycle's period, the propagation delay or time- difference-on-arrival (TDOA) of the single bounce and double bounce multipath interference terms must satisfy time coincidence criteria. The propagation factor (PF_uwb) derived in this paper averages the range and frequency variations of the derived gain/loss function G(f,R) into a single value which can be used as input to a UWB system loss factor for short range air defense (SRAD) scenarios.

This paper reveals a detection performance advantage of polarimetric Ultra Wide Band (UWB) systems over polarimetric Narrow Band (NB) systems. With UWB to resolve the target of interest into its Multiple Dominant Scatterers (MDSs), the so-called signal cancellation problem associated with polarimetric NB systems is essentially eliminated via randomization of MDS polarization, resulting in a much more reliable detection performance improvement than what NB systems can offer with polarization processing.

In recent years, Syracuse Research Corporation has implemented bandwidth expansion and extended coherent processing techniques to improve scatterer resolution of images derived from existing data. The algorithms can produce extraneous responses, however, which is not the case if the data is collected using an ultrawide bandwidth radar. The long and continuous effort in radar imaging at SRC and the limited number and availability of wideband sensors justified the development of an in-house imaging facility that would be available on demand. As a result, the SRC Ultrawide Bandwidth Measurements Radar was assembled to perform the required measurements of the radar cross section (RCS) of isolated scatterers and to determine their relative location on a wide variety of target types. Along with the use of radar images for target analysis has been the development of target modeling algorithms and software that accurately predict the electromagnetic scattering from complex objects. A few years ago, the software was expanded to model objects as viewed in wide angle bistatic configurations. The SRC radar was utilized to validate these modeling algorithms in several radar bands at large bistatic angles. The system has subsequently been employed to measure a variety of target models and complex shapes and to evaluate the effects of radar absorbing materials (RAMs).

The Transient Electromagnetic Scattering Laboratory (TESL) is described which employs a unique dual-channel ultra-wideband impulsive illuminating source. This is a free-field facility where targets are suspended within an anechoic chamber. A highly coherent primal step pulse is amplified by two GaAs wideband power amplifiers having complementary passbands which feed individual wideband horn antennas. This yields an effective 1 - 12 GHz impulse bandwidth. A high speed digital processing oscilloscope samples the output of a single receiving horn. The TESL has facilitated research into radar target identification using complex natural resonances. Theory and operational characteristics of the facility are discussed and technical improvements are described which have yielded significant improvement in both the effective bandwidth and the signal-to-noise ratio of transient scattering measurements. Experimental validations are shown which illustrate the level of fidelity attainable and consideration is given to recent enhancements, including an increase of measurement bandwidth to 50 GHz.

The response of ultra-wideband (UWB) signal acting on the coating radar absorbing material (RAM) targets have been investigated experimentally in this paper. The results of measurement and calculation are given. The difference between the responses of impulse signal acting on the metallic targets with RAM and without RAM have been compared. It is shown that the UWB signal superior 10 dB to the narrowband (NB) signal of conventional radar to combat the coating RAM's targets, and clearly indicated that UWB signals have surely good capabilities to combat the NB RAM.

The recent interest in the potential of ultra-wideband (UWB) radar techniques has given rise to many fundamental questions on the relative merits of UWB versus more traditional radar designs. SRI has been actively engaged in the use of UWB radar for a variety of applications, including ground-penetrating synthetic aperture radar. Synthetic aperture radar (SAR) techniques are a convenient way of obtaining meaningful images from radar return signals, and the UWB/SAR combination may prove to be a powerful tool for imaging buried mines and mine fields. This paper discusses the results of a project to use a UWB/SAR to image buried mines. The UWB/SAR technique is explained briefly and experimental results are presented.

A line segment image transform and inverse transform is used to detect and extract quasi- linear features in synthetic aperture radar (SAR) images. The transform is a windowed version of the Radon transform. The transform begins by dividing the original image into overlapped subimages using a simple analysis filter. A Radon transform is applied to each subimage yielding a representation of the subimage in terms of line segments at varying angles and positions. The amplitude of each line segment is used to calculate the detection statistic for that line segment. Line segments whose detection statistics are above a specified threshold are passed, and the filtered subimage is reconstructed using an inverse Radon transform (convolution back projection). Finally, the filtered subimages are recombined to form the filtered version of the original image. The filtering procedure passes quasi-linear features in the original image, and rejected features that are not quasi-linear, such as speckle. The inversion procedure is designed so that the original image is reconstructed if the threshold is set to zero.

The recognition of targets in synthetic aperture radar (SAR) imagery using a quadratic classifier is proposed. Correlators are used to compute distances under an optimum transform to measure similarity between ideal reference images and the actual data. The transform is a filter which responds to features specifically useful for discrimination. This is attractive for model based training since only a similarity in features is required between the actual images and their class models rather than a precise match in pixel values. The quadratic terms are unaffected by shifting of the input image while linear terms are computed using shift-invariant correlation. The system is thus non-linear but shift-invariant. Specifically, 'distance' vectors are generated by the filter banks which are analyzed by a rudimentary rule base to determine whether the input is a target image or clutter. In this paper, we describe a SAR automatic target recognizer (ATR) with results for 3 and 5 class problems. the data used is actual SAR imagery of military targets.

The signal processing for a down-looking airborne radar requires that the input signal data be compensated for platform motion and terrain variations. Typical examples are a Moving Target Indicator (MTI) function in a surveillance radar where the ground clutter must be maintained around zero Doppler in order to accurately detect moving targets, and a Synthetic Aperture Radar (SAR) where the correct location of the reflectivity of a scatterer depends on maintaining an accurate Doppler centroid for each pixel position. Platform motion compensation can be complicated by a wide elevation beam which intercepts many range cells at each azimuth, forcing the compensation of many range cells with the platform rates. These platform rates are assumed to be available from on-board inertial navigation system (INS), and the motion compensation correction can be obtained using the INS inputs. Processing of ground returns will be further complicated by the lack of reliable terrain information which can introduce significant Doppler errors into the motion compensation. Such errors will have an effect on the motion compensation, hence on the positioning of ground returns in the Doppler domain. Variations due to local topography are impossible to model accurately without prior knowledge of the terrain.

The Environmental Research Institute of Michigan (ERIM) has developed a unique ground- based, portable, synthetic aperture radar (SAR). This SAR images targets in their natural backgrounds without the expense of an airborne sensor and with higher performance (bandwidth, resolution) than existing airborne systems. A horizontal 36-foot long aluminum truss supports a rail and an antenna carriage, which is moved along the rail to allow synthetic aperture focusing. The system is fully-polarimetric and has collected data over the frequency band of 400 - 1300 MHz resulting in a nominal resolution of 0.17 m in range and 0.5 m in cross-range. Because the system is ground-based and images are in the near field, a traditional spotlight SAR processor is inadequate. ERIM has implemented a near-field SAR processor which transforms the data to the far-field by performing a Fourier transform in the azimuth direction. This is often referred to as a planewave decomposition and results in data sampled as if it were collected in the far-field. The final step is a traditional spotlight processor (resampling and two-dimensional Fourier transform) which is applied to the data to form the image.

As part of an ongoing effort to determine the utility of Ultra-Wideband radar systems for military application, an experimental program was developed and executed to collect terrain clutter data using high resolution waveforms in the UHF spectral region. Two approaches to the design of the radar instrumentation to be used to collect this data were considered: an impulse system with a nominal 1 nsec pulse duration, and a 'conventional' stepped chirp instrumentation radar covering the same frequency range. A novel feature of the program was the use of a scanned linear aperture to simulate the use of a large, ideally-weighted real aperture antenna system. In this paper we report the theoretical analysis done to predict and compare the performance expected from either system approach. This analysis is presented in terms of the noise-equivalent reflectivity of the clutter measurement system, time to collect data, and the impact of the linearly scanned aperture on sensitivity, angular resolution, and data collection time.

This paper reports the results of ultrawideband radar clutter measurements made by Battelle- Pacific Northwest Laboratories and the System Planning Corporation near Sequim, WA. The measurement area is a mountainous coniferous forest with occasional roads and clear-cut areas. Local grazing angles range from near zero to approximately 40 degree(s). Very limited data are also presented from measurements made in a desert-type terrain near Richland, WA. Two ultrawideband radar systems were employed in the data collection. An impulse system providing an approximate one nanosecond monocycle pulse (bandwidth of 300 MHz - 1000 MHz) acquired data over a 0.7 km² area (121,000 resolution cells). A step chirp radar with the same total bandwidth as the impulse system collected data over a 6.2 km² area (780,000 resolution cells), including the area sampled by the impulse system. Wideband TEM horn antennas (log-periodic antennas for the step chirp system) deployed on a 19 m horizontally scanned aperture were used for transmission and reception, providing a 1.5 degree(s) azimuth resolution at 300 MHz for both systems.

Array-antenna patterns are calculated for arrays of radiators of time-limited signals. Antenna patterns are calculated in the time domain by superposition of temporal waveforms at various angles. Monocycle waveforms are used as canonical building blocks to illustrate the method. Pulses are short compared to the dimensions of the antenna and interference phenomena are localized to a percentage of the total aperture. Low sidelobes are achieved through two methods: (1) the limited interference phenomena afforded by monocycle excitation, and (2) randomization of individual element placement. Randomization of element placement serves to destroy periodicities in the physical aperture. In rectangular arrays where elements are arranged in rows, periodicities in the lattice along the principal planes make sidelobes higher along these planes than for off-axis directions. This behavior is evident in arrays using both short-pulse and sinusoidal excitation. Periodicities in the physical aperture of arrays of short- pulse radiators have also been associated with a narrow-banding effect manifesting itself as spikes in the spectral response at particular angles. Randomizing the positions of the individual radiators, each radiating a broad-band signal, helps keep the spectrum broad for all angles. Sidelobe levels have been demonstrated which are well below the 1/n limit associated with unipolar-impulsive signals where n is the number of radiating elements.

In the paper, synthetic aperture radar imaging of objects imbedded in a half space lossy medium at close range is described. A general signal processing technique is presented which can obtain one-, two-, and three-dimensional high resolution images of a target imbedded in a half space lossy medium from data collected at very close range to the aperture. The technique is based on the matched filter concept (or focusing function). It used tomographic reconstruction algorithm for two- and three-dimensional imaging. A general expression of resolutions of the imaging system is given.

The specific source requirements of UWB Radars have been difficult to meet with existing microwave sources. BASS provides a technology which makes long range, high resolution UWB Radars feasible. The high peak power and burst rate capability of BASS-based microwave generators when combined with their exceptional reliability, temporal, amplitude and spectral stability results in an ideal source for UWB Radar applications. We demonstrate the unique capabilities of the BASS technology in the following discussion.