Algorithms for estimating the angular velocity, angular acceleration, orientation, length, and width of an ISAR target and for identifying the scattering points of ISAR targets are derived and demonstrated. A correspondence between the ISAR motion estimation operations and a rigid body dynamical model is illustrated. It will be shown that an ISAR image amounts to projecting scatterers into a plane perpendicular to the target axis of rotation and coplanar with the range axis. The proper Doppler scale factor of imagery and its variation is proportional to the rotation rate and acceleration. These result as a by-product of SAIC ISAR motion estimation techniques. Improved geometric estimates can then be found by incorporating modeling assumptions and a priori knowledge. A number of land and sea examples are provided to reinforce the above concepts.

In ISAR imaging, a Point Scatterer mechanism is frequently invoked, implicitly or explicitly. This paper looks at which categories of EM scattering give rise, in effect, to point scatterer type of return. We find that there are a number of different types of point scattering returns. Such an examination is also relevant to the adequacy of point scatterer models (PSM) which have sometimes been used for modeling aircraft or ships. Another important question is how point scatterers can be recognized as such. Finally the impact of point scatterer assumptions on ISAR autofocus methods will be briefly assessed.

An important question concerning the practical feasibility of radar target identification is whether specific scatterers on a target can be observed over relatively large aspect angle sectors. Using turntable data on a fighter aircraft, we demonstrate that this is indeed the case. We show that essentially the same scatterers can be observed when the aspect angle is changed in steps from 5 degrees to 70 degrees. Although some scatterers on the fuselage cannot be observed at the larger aspect angles, this is due to the fact that the data allowed implementing a crossrange resolution of only 4 ft. This implies a lack in resolution performance along the fuselage as the aspect angle becomes larger. The loss of some scatterers at larger aspect angles thus is due to poor resolution rather than a lack in scatterer persistence.

The 'weak scatterer' approximation is usually inappropriate for re-entrant target structures, and ISAR images based on this approximation often display unwanted 'artifacts' which can complicate image interpretation. Last year, we presented an image restoration method which can be used to mitigate these artifacts without significant impact on neighboring image components and demonstrated its application using anechoic chamber data [SPIE Proc., Radar Processing, Technology, and Applications II, vol. 3161, pp. 9 - 19 (1997)]. Since then, we have further examined this novel filtering technique and applied it to measured ISAR data. In addition, duct shape-specific parameters can be recovered using this analysis, and we have looked at the possibility of applying these to the problem of radar-based target classification. Our presentation will review the past theory and present results of our investigations over the past year.

Standard ISAR image focusing techniques utilize the magnitude and position of range domain target signatures to achieve translational motion compensation. Alternatively, range can be determined by analyzing the slope of the unwrapped phase function associated with the frequency domain signature of a moving target. We propose a motion compensation method based on such phase slope analysis. To improve noise performance, the probability density function of additive noise corrupting the signature is modeled as a Weibull distribution to determine a threshold for constant false alarm rate filtering. Complex analysis is then utilized to determine sample intervals where noise is relatively weak, where the amplitude of the filtered signature is high above the noise interference level and the phase of the same signature is nearly linear Weighted least squares is subsequently used to estimate the target's kinematic parameters based on the phase slope measurement. Results show that parameter estimates for motion compensation obtained in this manner converge to a unique solution, thus providing focused ISAR imagery of the target.

According to the different problems and techniques related to the detection and recognition of airplanes and vehicles moving on the Airport surface, the present work mainly deals with the processing of images gathered by a high-resolution radar sensor. The radar images used to test the investigated algorithms are relative to sequence of images obtained in some field experiments carried out by the Electronic Engineering Department of the University of Florence. The radar is the Ka band radar operating in the'Leonardo da Vinci' Airport in Fiumicino (Rome). The images obtained from the radar scan converter are digitized and putted in x, y, (pixel) co- ordinates. For a correct matching of the images, these are corrected in true geometrical co-ordinates (meters) on the basis of fixed points on an airport map. Correlating the airplane 2-D multipoint template with actual radar images, the value of the signal in the points involved in the template can be extracted. Results for a lot of observation show a typical response for the main section of the fuselage and the wings. For the fuselage, the back-scattered echo is low at the prow, became larger near the center on the aircraft and than it decrease again toward the tail. For the wings the signal is growing with a pretty regular slope from the fuselage to the tips, where the signal is the strongest.

This paper addresses the problem of resolving targets in a Synthetic Aperture Radar (SAR) scene. The approach is based on computing an analyzing the SAR signature of a localized or spotlighted target area in the reconstructed SAR image at a fixed fast-time frequency (single tone) of the radar signal. This signature is translated into a set of fringe patterns in the spatial domain of the SAR image; the fringe patterns is used to identify the presence of more than one target in the spotlighted target area. The resolution of this scheme is shown to surpass the conventional resolution of a SAR system. Results using the realistic turntable data of a tank and a set of corner reflectors are provided.

The University of Nebraska has developed an ultrawideband coherent random noise radar that accomplishes phase-coherent processing of the received data. The system operates over the 1 - 2 GHz frequency range, and achieves phase coherence using heterodyne correlation of the received signal with the time delayed replica of the transmitted signal. Knowledge of the phase of the received signal and its time dependence due to the motion of the target permits the system to be configured as a Doppler radar for detecting both linear and rotational motion. Preliminary simulation and experimental results presented last year indicate confidence in the system's ability to extract linear and rotation Doppler velocities of targets. The accuracy with which Doppler spectra of moving objects can be estimated is dependent not only upon the phase performance of various components within the radar system, but also upon the uncertainties arising from random and systematic internal and external factors. This-paper describes the simulation studies to characterize the uncertainties in Doppler measurement due internal and external mechanism.

The University of Nebraska has developed an ultra-wideband (UWB) coherent random noise radar that accomplishes phase- coherent processing of the received data. The system operates over the 1 - 2 GHz frequency range and achieves phase coherence using heterodyne correlation of the received signal with the time delayed replica of the transmitted signal. The system coherence allows for extraction of a targets polarimetric amplitude and phase characteristics. Collecting data from a rotating target over a series of range bins may be interpreted as construction of projections of a targets reflectivity function. This paper describes ISAR imaging with this system using tomographic methods. This paper gives a brief overview of the theory of image reconstruction from projections, the theory of random noise radar polarimetry, and presents simulations and initial experimental results.

FM-CW radars operating in the millimeter wave or upper microwave bands can provide low cost, low power solutions for many applications requiring the resolution of targets separated by one meter or less in range. Range resolution of this quality is obtained by sweeping the radar output frequency over several hundred megahertz of bandwidth using modern techniques to achieve extremely good linearity. Because of the short wavelengths at millimeter bands, relatively good angular resolution is achievable with moderately sized antennas. Applications for FM-CW radar sensors include automotive collision warning systems, traffic monitoring, height profiling, terrain profiling, autonomous vehicle navigation, surveillance and site security systems where high resolution is required.

Monostatic and bistatic Synthetic Aperture Radar (SAR) imaging systems with Wide-Bandwidth Continuous-Wave (WB-CW) sources have been utilized for military reconnaissance. WB-CW sources are less susceptible than FM-CW sources to Electronic Counter Measures (ECM). The main shortcoming of the WB-CW microwave illumination is that its resultant SAR echoed signal is not composed of distinct Doppler spreading around specific tones; this creates difficulties to formulate the image formation in the WB-CW SAR systems via the conventional pulse or FM-CW SAR imaging algorithms. The current paper outlines a Time Domain Correlation (TDC) processing method and a Fourier-based processing method for image formation in WB-CW monostatic and bistatic SAR systems. Results are provided.

Point spread functions produced by three families of curvilinear apertures are derived and their characteristics evaluated for three-dimensional imaging. Outputs have been obtained for previously proposed speedups in the PML algorithm and consideration for further refinement introduced. Preliminary autofocus results making use of the stability afforded by the PML algorithm are presented.

Traditional medium and high PRF Pulse Doppler waveforms suffer from range ambiguities caused by the relatively high PRF. These ambiguities also translate into higher dynamic range requirements which result from having clutter from many Pulse Repetition Intervals (PRI) summed at RF. In order to adequately process out the clutter, it is imperative that the receiver possess enough dynamic range to allow clutter vectors from all range intervals to sum vectorally without any distortion. In addition, it is imperative to insure that any significant clutter is represented in every PRI processed, which usually translates into adding fill pulses to the dwell or processing less pulses in the coherent processor. Wideband step frequency waveforms, traditionally used to obtain high range resolution, have some interesting properties that mitigate some of these problems normally associated with Pulse Doppler waveforms. This paper discusses the use of a wideband stepped frequency waveform for reducing dynamic range, providing some selective range filtering, and reducing the processing speed requirements imposed on the A/D converter and subsequent processing.

Tracking a low-altitude target in elevation is difficult when both direct and reflected radar returns originate within the main beam of the interrogating radar. In such conditions, a conventional monopulse radar is subjected to increased angle noise and bias error in elevation. This is a long-standing, unsolved (in a practical sense) problem in low-elevation target tracking. A novel application of a target extent estimator ('C²') has recently been shown theoretically capable of mitigating both specular and diffuse multipath interference in low-elevation tracking situations. Under Navy Phase II SBIR funding, ORINCON, Signalogic and TRW have collaboratively investigated the application of this technology by processing data collected during X-band experiments on real targets in low-level flight at NAWCWPNS, Point Mugu in early 1998. We describe the design and performance of these flight test experiments and summarize the salient off-line signal processing results, which demonstrated reliable elevation tracking in the horizon region with rms errors typically on the order of one-twentieth of a beamwidth. We conclude with a discussion of areas for future research and development of this technology.

We describe a jamming cancellation algorithm for wide-band imaging radar. After reviewing high range resolution imaging principle, several key factors affecting jamming cancellation performances, such as the 'instantaneous narrow-band' assumption, bandwidth, de-chirped interference, are formulated and analyzed. Some numerical simulation results, using a hypothetical phased array radar and synthetic point targets, are presented. The results demonstrated the effectiveness of the proposed algorithm.

A new methodology based on adaptive joint time-frequency processing is proposed to separate the interference due to fast rotating parts from the original ISAR image of the target. The technique entails adaptively searching for the linear chirp bases which best represent the time-frequency behavior of the signal and fully parameterizing the signal with these basis functions. The signal components due to the fast rotating part are considered to be associated with those chirp bases having large displacement and slope parameters. While the signal components due to the target body motion are represented by those chirp bases having relatively small displacement and slope parameters. By sorting these chirp bases according to their slopes and displacements, the scattering due to the fast rotating part can be separated from that due to the target body. Consequently, the image artifacts overlapping with the original image of the target can be removed and a clean ISAR image can be produced. Furthermore, useful rotation rate information contained in the Doppler signal can be extracted. Successful applications of the algorithm to numerically simulated and measurement data show the robustness of the algorithm.

Chen recently presented an ISAR imaging technique using the joint time-frequency analysis (JTFA), which has been shown having a better performance for maneuvering targets over the conventional Fourier transform method. It is because the frequencies of the radar returns of the maneuvering targets are time-varying and JTFA is a technique that is suitable for such signals. It is also known that JTFA concentrates a signal, such as a chirp, while spreads noise. In this paper, we study the signal-to-noise ratio (SNR) in the ISAR imaging using the JTFA. We show that the SNR increases in the joint TF domain over the one in the time or the frequency domain alone both theoretically and numerically. This shows another advantage of the JTFA technique for the ISAR imaging.

An approach is presented for the design of time-varying filters that separate selected ISAR target signatures with a constant false alarm rate. Here, the time-frequency representation of each signature is utilized to define a binary filter which is enhanced and labeled using computer vision techniques. Upon filtering, the signatures of the targets are synthesized from their Short Time Fourier Transform representations. In order to obtain a focused radar image for each target, its filtered signature must be motion compensated in the frequency domain. Therefore, radar images can be generated for scenarios in which multiple moving targets can not be imaged otherwise. As an instance, the response of a stepped frequency radar system is simulated to demonstrate that combined signatures of aircraft are effectively separated, compensated and processed into individual images.

The concept of spatial time-frequency distribution (STFD) has been recently introduced and applied to solve the problem of blind source separation for nonstationary signals. In this paper, we propose to apply this same concept to solve the problem of the direction of arrival (DOA) estimation. A new method for the estimation of the signal subspace and noise subspace based on time-frequency signal representations is introduced. The proposed approach consists of the join block- diagonalization (JBD) of a combined set of spatial time frequency distribution matrices. Once the signal and the noise subspaces are estimated, any subspace based approach can be applied for DOA estimation. Herein, we propose to use the MUSIC algorithm. Performance comparison of the proposed approach with the classical MUSIC method is provided. The influence of the time-frequency kernels on the performance of the newly proposed Time-Frequency MUSIC (TF-MUSIC) is evaluated numerically.

We present the analytical description of a photorefractive phased array beamforming system using the BEAMTAP (Broadband and Efficient Adaptive Method for True-Time-Delay Array Processing) algorithm for a large N-element array that requires only 2 tapped delay lines (TDLs) instead of the conventional N TDLs. Simulation results indicate that the processor is able to adapt to a broadband signal of interest at a specific angle of arrival. We show that the system produces a coherent sum of the desired signals from the phased array, with the corresponding time delays appropriately compensated for in an adaptive fashion without prior knowledge of the angle-of-arrival.

In this paper we consider a discretized version of the problem of optimal beam-forming, or radar transmit and receive pattern design, for stationary radar target localization in the presence of white Gaussian noise. We assume that the target is equally likely to be in one of N discrete cells and the number of allowed observations L is strictly less than N, making an exhaustive search not feasible. We propose two new approaches for beam-form design in target localization problems: a fixed, off-line beam-form design approach and an adaptive, on-line beam-form design technique. The beam-form is designed off-line in the fixed approach to minimize the probability of error after exactly L observations. In particular, the decision is available only after the last (L^the) observation is acquired. We show that this fixed beam-form design approach is directly related to signal constellations design in digital communications. By contrast, the beam-form is optimized after each observation to minimize the probability of incorrectly localizing the target after the next observation is acquired at each step of the process. The optimization relies on the previously acquired information. The adaptive approach has a better performance than the fixed one. Unlike binary search, these two approaches can work with any number of observations. This work falls under the area of optimal search, which deals with optimal allocation of effort in search problems. The need for optimal search strategies arises in many areas such as the radar target localization problems that we address here, fault location in circuits, localization of mobile stations in wireless networks and Internet information searches.

Invariant features in two dimensional binary images are extracted in a single layer network of locally coupled spiking (pulsating) model neurons with prescribed synapto-dendritic response. The feature vector for an image is represented as invariant structure in the aggregate histogram of interspike intervals obtained by computing time intervals between successive spikes produced from each neuron over a given period of time and combining such intervals from all neurons in the network into a histogram. Simulation results show that the feature vectors are more pattern-specific and invariant under translation, rotation, and change in scale or intensity than achieved in earlier work. We also describe an application of such networks to segmentation of line (edge-enhanced or silhouette) images. The biomorphic spiking network's capabilities in segmentation and invariant feature extraction may prove to be, when they are combined, valuable in Automated Target Recognition (ATR) and other automated object recognition systems.

In the naval electronic environment, pulses emitted by radars are collected by electronic support measures receivers. The aim is to gather these pulses in such way that one cluster corresponds to one radar despite the waveform parameters agility. To achieve it, this paper describes a pulse train deinterleaving process using a multi-hypotheses architecture. A hypothesis tree, built from pulse measurements, represents all the possibilities to associate pulses to emitters, according to one hypothesis of pulses association. Different processes are used to find the valid hypothesis that correspond to emitters. These processes have two aims: avoiding the combinatorial explosion of the number of hypotheses, and to lead to the solution tree: one branch corresponds to one effective emitter. Radars waveforms with agile parameters are considered. A new process has to be added to take into account missing pulses.

For control of space debris of both artificial and natural origin, the passive ground based on the optical system is proposed consisting of three consequent and interconnected channels: (1) the forming channel on basis on an aperture synthesis array; (2) the detecting channel on the basis on photon counting mode in the image; (3) the processing channel on the basis of algorithmic software of digital correlational images processing. The principles of construction and application strategy of the proposed system are considered.

For the problem of estimating the difference in arrival times of radio waves impinging on two spatially separate antennas for the purpose of passively locating the source of a communications or telemetry signal, a new parameter estimation algorithm that is highly tolerant to interference and noise is proposed. By exploiting the cyclostationarity property of the signal of interest, the new algorithm which models the time delay as a finite impulse response filter removes the effects of additive noise and interfering signals. Then from a set of linear equations involving cyclic (cross) correlation, the time delay can be estimated to be the index of the FIR parameter which has maximum value. The new approach exhibits its signal selectivity regardless of the extent of temporal, spectral, or spatial overlap among received signals, it is only required that the signal of interest has a known (or measurable) analog carrier frequency or digital keying rate that is distinct from those of all interfering signals. Yet the computational complexity of this new approach is comparable to that of conventional methods.

The method of matched pursuits is used to analyze two- dimensional digitized SAR images generated by XPATCH in the digital spatial domain. Several algorithms for employing matched pursuits to the problem of detection and recognition of targets in the digital spatial domain are presented. These algorithms are then demonstrated through application to SAR images of four American military aircraft: the F-15, the T-38, the V-218, and the X-29. Several chip dictionaries are proposed and evaluated. A performance comparison between the method of matched pursuits and the M-ary optimum detector is provided. The degradation of performance in the presence of noise and computational complexity issues are addressed.

The University of Nebraska has developed an ultra-wideband coherent random noise radar that accomplishes phase-coherent processing of the received data. The system operates over the 1 - 2 GHz frequency range. In order to make calibrated radar cross section measurements of targets and terrain, a radar calibration target was fabricated and tested. The unique requirements for the ultra-wideband calibration target include (1) high radar cross section value to minimize effects of background reflections, (2) constant radar cross section over the frequency range to ensure calibration accuracy, and (3) wide beamwidth to minimize effects of antenna pointing errors. The design consisted of a receive and a re-transmit antenna between which a high-pass filter and a microwave amplifier were inserted. Log-periodic antennas were used as calibration target antennas owing to their broadband and wide beamwidth characteristics. The high-pass filter possessed a 12 dB per octave roll off to appropriately reduce the signal level at lower frequencies to compensate for the correspondingly lower propagation loss as predicted by Friss transmission formula. The high-gain broadband amplifier was used to provide a high- retransmitted power level back to the radar. The design and performance characteristics of the active ultra-wideband radar calibration target are discussed in this paper.

Intelligent vehicles of the future will be guided by radars and other sensors to avoid obstacles. When multiple vehicles move simultaneously in autonomous navigational mode, mutual interference among car radars becomes a serious problem. An obstacle is illuminated with electromagnetic pulses from several radars. The signal at a radar receiver is actually a mixture of the self-reflection and the reflection of interfering pulses emitted by others. When standardized pulse- type radars are employed on vehicles for obstacle avoidance and so self-pulse and interfering pulses have identical pulse repetition interval, this SI (synchronous Interference) is very difficult to separate from the true reflection. We present a method of suppressing such a synchronous interference. By controlling the pulse emission of a radar in a binary orthogonal ON, OFF pattern, the true self-reflection can be separated from the false one. Two range maps are generated, TRM (true-reflection map) and SIM (synchronous- interference map). TRM is updated for every ON interval and SIM is updated for every OFF interval of the self-radar. SIM represents the SI of interfering radars while TRM keeps a record of a mixture of the true self-reflection and SI. Hence the true obstacles can be identified by the set subtraction operation. The performance of the proposed method is compared with that of the conventional M of N method. Bayesian analysis shows that the probability of false alarm is improved by order of 10³ to approximately 10⁶ while the deterioration in the probability of detection is negligible.

Synthetic Aperture Radar utilizes the motion of the platform on which it is mounted to realize a virtual array antenna. The illumination of a specific patch of ground and the sampling of the radar returns over time and space result in a non-uniform collection space. Motion compensation corrects for certain platform motions, but still may result in a polar collection space with non-uniformly spaced samples. To employ the fast Fourier transform to form an image from this phase history data, the polar raster data must be resampled onto a rectangular grid by means of polar reformatting. However, non- uniformity errors present in the polar grid will propagate to the rectangular grid, and result in SAR image defocusing. This paper describes a technique that may be used to interpolate the non-uniformly collected radar samples to a uniform polar space, thereby improving the efficacy of the polar-to- rectangular resampling process. This work presents new ideas for polar grid design by demonstrating that certain virtual grids are better for interpolation than others. A measure of the relative interpolation error is established. The step-by- step approach of polar grid interpolation is explained.

This project assembled a SAR algorithm analysis testbed and demonstrated the ability to focus moving targets possessing non-linear motion over a synthetic aperture collection. The testbed consisted of the ImSyn optoelectronic processor connected to a low end SGI workstation. A suite of software tools were developed to support focusing moving targets and to provide the necessary user interaction with the imaging process. We developed and tested moving target algorithms on a variety of simulated data sets and real SAR data. This report details the approach and algorithms that were employed to successfully process both SAR and ISAR data collections to add value to target discrimination and non-cooperative target identification.

Inverse synthetic aperture radar (ISAR) imaging on a turntable-tower test range permits convenient generation of high resolution two-dimensional images of radar targets under controlled conditions for testing SAR image processing and for supporting automatic target recognition (ATR) algorithm development. However, turntable ISAR images are often obtained under near-field geometries and hence may suffer geometric distortions not present in airborne SAR images. In this paper, turntable data collected at Georgia Tech's Electromagnetic Test Facility are used to begin to assess the utility of two- dimensional ISAR imaging algorithms in forming images to support ATR development. The imaging algorithms considered include a simple 2D discrete Fourier transform (DFT), a 2-D DFT with geometric correction based on image domain resampling, and a computationally-intensive geometric matched filter solution. Images formed with the various algorithms are used to develop ATR templates, which are then compared with an eye toward utilization in an ATR algorithm.

Resolution is a fundamental limitation of any processing based on radar data. Conventional radar imaging techniques, in general, make use of the FFT to determine the spatial location of a target from its scattered field. The resolution of these images is limited by the bandwidth of the interrogating radar system and the aspect angle sector over which the target is observed. In such cases, superresolution offers the potential to improve system performance by increasing the resolution. Superresolution is the process of increasing the effective bandwidth of an image (or time series) by introducing collateral data to augment the dataset; thus the Rayleigh resolution imposed by the size of the dataset is overcome by the introduction of the synthetic collateral data. This paper presents a state of the art survey of radar superresolution, applicable to both 1-D and 2-D data and a comparison of superresolution algorithms using real and simulated data sets. Complex data sets are chosen so as to mimic scenes with a large number of scattering mechanisms. The paper also presents specific applications of superresolution for air-to-ground surveillance, data resolution enhancement, SAR ATR and FOPEN ATR.

As a target moves in the cross-range (azimuth) direction, it rotates relative to a radar. Over a limited observation interval, target points at different azimuths have different range rates that are approximately constant with time. The discrete Fourier transform can be used to construct multi- pulse filters that are matched to the different constant- frequency Doppler components. Maneuvers and other unpredictable effects introduce time varying range rates that defocus the target image on a range-Doppler map. Representations of instantaneous frequency vs. time attempt to reduce blurring by accurate portrayal of time-varying Doppler shifts. An alternative, ideal receiver (correlation) approach to adaptive ISAR focusing is considered here. The predicted delay history of each target point is corrected so as to maintain focus when the corresponding reference function is correlated with echo data. Proposed delay and/or target rotation corrections can be evaluated by comparing test images. For a delay-and-sum synthetic beam former, the n^th test image is formed by adding delay-corrected samples of the n^th echo to an image that has been sequentially constructed with previous echoes. Image bandwidth can be used as a focus measure for selecting the test image with the best delay/rotation correction. The final image is sequentially constructed from the best test images. The resulting image- based tracker can incorporate a dynamic model as in Kalman filtering and is similar to time warp compensation in speech classifiers.

This paper investigates methods to improve template-based synthetic aperture radar (SAR) automatic target recognition (ATR). The approach utilizes clustering methods motivated from the vector quantization (VQ) literature to search for templates that best represent the signature variability of target chips. The ATR performance using these new templates are compared to the performance using standard templates. For baseline SAR ATR, the templates are generated over uniform angular bins in the pose space. A merge method is able to generate templates that provide a nonuniform sampling of the pose space, and the templates produce modest gains in ATR performance over standard templates.

Small random rotational motion of an aircraft can affect the fidelity of the target's High Range Resolution (HRR) profile. A multiple scattering point-source model has been developed to investigate the distorting effect on a simulated target. Results indicate that even when the target possesses a very small amount of random motion during radar interrogation using a stepped frequency waveform (SFWF) scan, sizeable distortion in the target's range profile can still occur, making target identification more difficult. The well known range walk effect of the target during the SFWF radar scan offers a partial explanation for the distortion of the target's range profile. A more interesting situation emerges when the rotational motion is time-varying; a more severe distortion can occur as a result, leading to a broadening phenomenon and spurious peaks appearing in the target's range profile.

Essex has developed a hybrid acousto-optic/digital electronic processor, called the Advanced Optical Processor (AOP) a.k.a. Wideband Range-Doppler Imager (WRDI), that generates high dynamic range, high resolution range-Doppler images from wideband radar returns. This processor supports high resolution processing necessary for target discrimination and kill assessment. The processor is described and results of testing with simulated and real field data are presented. Key capabilities of this processor are: (1) high dynamic range to detect small cross section targets in a severe clutter background, (2) large image sizes in multiples of 1024 range bins by up to 128 Doppler resolution bins, (3) dynamically adjustable Doppler resolution, (4) dynamic reconfigurability of modules to switch between coarse range resolution covering a large range extent for acquisition mode to a fine resolution mode for target discrimination, (5) the ability to accommodate time compression/dilation to eliminate blurring of moving targets in high resolution range-Doppler images, and (6) the ability to efficiently combine multiple low bandwidth, low range resolution radar returns to obtain high range resolution range-Doppler images. The AOP can process true arbitrary waveforms, which are needed to support the high dynamic range required for discrimination. This is achieved in a compact, light weight and cost effective package.