The waveform and required signal processing for a new multiband shipboard radar is described and some preliminary test results are presented. The waveform is a high PRF stepped frequency waveform designed to provide coherent gain in the order of 40 dB, clutter rejection in the order of 100 dB, 1 foot range resolution, and single dwell unambiguous range and velocity measurements.

This article addresses the use of the Fourier transform receiver as a radar system for high-resolution range-profile generation. A significant point of reference is the equivalence of this receiver to the standard matched filter. Henceforth, it is demonstrated that range information can be extracted from beat-frequency measurements by processing a selected set of waveforms. Attention is given to the step and hop-frequency modulation waveforms, which are shown to be easily compressed into an optimum range profile via coherent, narrowband processing. For these waveforms, the effects of velocity are also discussed in the context of profile distortion due to Doppler mismatch.

Ultra-wideband radar systems provide great potential for radar target detection, identification and imaging through their inherent high-resolution capabilities. This paper considers two applications of a stepped-frequency ultra- wideband radar--detection of targets close to a disturbed sea surface, and imaging of airborne targets. A new technique for target detection is presented, based on the E- pulse concept and designed to eradicate the sea clutter signal while enhancing the target response. A simulation of a missile travelling above an evolving sea-water model is considered, and results are compared to measurements made in an anechoic chamber. Finally, the effects of signal bandwidth and bistatic angle on image resolution are explored, using a time-domain imaging identity with measured, band-limited signals.

Based on the statistical analysis of the calculated glint of an aircraft and the measuring glint, the long term dependence of target glint is shown in this paper. Due to the similar correlation structure and spectrum structure between target glint and the filtered discrete fractionally Gaussian noise process, a new idea, that glint can be modeled by the filtered discrete fractionally differenced Gaussian noise in statistical sense, is developed in this paper. This model provides a simple and valid way to simulate target glint due to the convenience of producing the filtered discrete fractionally differenced Gaussian noise sequence. Furthermore, from the point of view of this model, the effectiveness of available methods and possible methods to reduce target glint is discussed in this paper.

The performance of signal/image processing algorithms used to form radar images and identify targets depends on propagation effects such as time-varying multipath, dispersion, attenuation, etc. In this paper, the effects of multipath propagation conditions that result from terrain or man-made environments and noise are modeled and simulated to determine their effects on radar target signatures. Here, we calculate the radar signatures for a ground vehicle subject to the deleterious effects of both multipath and additive noise. A minimum mean-square-error Wiener equalizing filter is developed and applied to the distorted radar target signatures. The mean-square-error is calculated as a function of signal-to-noise ratio and compared for several cases to demonstrate the potential for improvement in radar image resolution. These techniques are particularly relevant for targets located in mountain or urban canyon environments.

In this paper we describe a method of obtaining nulls in a phased array system using sub-array based phase-only nulling concepts. Phase-only nulling in phased-array antennas is appealing because the phase shifters can be employed for the dual purpose of beam-steering and nulling of unwanted interference such as jamming, radio frequency interference and clutter. The technique is applicable to transmit as well as receiver nulling in the situation when the amplitude control is not available. The application here is on transmit nulling for jammers decoying and surface clutter suppression in the upper beams. Instead of phase of the entire array, a small set of sub-arrays on the periphery of the array are considered. This implementation allows minimal modification of the phase control structure of the phased array system. Phase-only nulling also has an obvious advantage for transmitting nulling if we want to weight uniformly for maximum output. The problem is solved by formulating it as an optimization problem with side constraints. The method is quite general and can be applied to symmetric as well as non-symmetric arrays with real or complex beams. Based on the analysis the system is then designed and results illustrated for a case similar to the proposed advanced Aegis system. The system simply consists of sub arrays which can be part of the main array clustered together, each cluster receiving only a single phase-change instruction for nulling in presence of jamming, in real time, involving very minimal computation, once the location of jammers are determined. The null-depths achieved shows a performance of about 50 dB's or more. Presently there are many radars which can b represented as circular arrays, with minor modifications, can be retrofitted with hardware for nulling in presence of multiple jammers or to suppress clutter.

Radar target identification requires the processing of a large amount of data. Radar cross section alone is a function of time/frequency, aspect angle, and polarization. Storing all this data as well as processing it for identification is an extremely difficult task. The goal of this research is to analyze the RCS of simple targets in the time/frequency domain to identify which time/frequency algorithms best highlight distinguishing target characteristics. We have taken extensive RCS measurements on simple objects, such as spheres, cubes, pyramids and corner reflectors, and processed the data using Fourier and wavelet transforms. In most instances, the data can be significantly compressed.

A system is proposed for joint tracking and recognition of airborne targets from the observations of radar sensors. It is assumed that the data available for the estimation of target orientation and recognition include sequences of range-profiles from a high resolution radar. Inference is performed using the posterior distribution on the complete parameters space, which includes the number of targets as well as their positions, orientations, and target types. The algorithm is critically dependent on appropriate sensor and target models, in the form of a likelihood for the range profiles given the target orientations, and a prior on the orientations determined by the target dynamics. Deterministic and stochastic models for high resolution radar data are presented, and the likelihood function under the deterministic model is examined. The viability of our approach is demonstrated through simulations that address two simplified recognition scenarios. The first simulation investigates joint tracking and recognition of a single maneuvering target from the simulated observations of both a cross-array tracking radar and a high resolution radar. In the second simulation, orientation estimation and recognition are performed for a single target which is approaching an airborne radar platform. Results from these simulations showing performance are given.

WISIP: wideband interferometric sensing and imaging polarimetry has become an important, indispensable tool in wide area military surveillance and global environment monitoring of the terrestrial and planetary covers. It enables dynamic, real-time optimal feature extraction of significant characteristics of desirable targets and/or target sections with simultaneous suppression of undesirable background clutter and propagation path speckle at hitherto unknown clarity and never before achieved quality. WISIP may be adopted to the detection, recognition and identification (DRI) of any stationary, moving or vibrating target or distributed scatterer segments versus arbitrary stationary, dynamically changing and/or moving geo- physical/ecological environments, provided the instantaneous 2 by 2 phasor and 4 by 4 power density matrices for forward- propagation/backward-scattering, respectively, can be measured with sufficient accuracy. For example, the DRI of stealthy, dynamically moving and/or camouflaged stationary objects occluded deeply into heterogeneous stationary and/or camouflaged stationary inhomogeneous volumetric scatter environments such as precipitation scatter, the ocean sea/lake surface boundary layers, the littoral coastal surf zones, pack-ice and snow or vegetative canopies, dry sands and soils, etc., can now be successfully realized. A comprehensive overview is presented on how these modern high resolution/precision, complete polarimetric co-registered signature sensing and imaging techniques, complemented by full integration of novel navigational electronic tools, such as DGPS, will advance electromagnetic vector weave sensing and imaging towards the limits of physical realizability. Various examples utilizing most recent image data take sets of the NAWC/ERIM-P3-UWB-TOPIF'E-CATI/LTBL- POLSAR and NASA-JPL-AIRSAR airborne, the NASA/DARA/DASI-SIR- C/X-SAR shuttle, and the ESA ERS-1/2 satellite imaging systems will be presented for demonstrating the utility of WISIP.

When generating inverse synthetic aperture radar (ISAR) images of a rapidly moving target such as an aircraft, motion compensation is an essential step. This process of removing the effects of the target's radial motion is required because the target usually moves a distance that is many times larger than the radar's range resolution during the time that the data for an image is being collected. This processing requires estimation of the target motion to within a fraction of a wavelength before useful images can be generated. To generate ISAR images in real- time, the estimation and correction processing must be performed hundreds of times per second. This level of processing performance is becoming possible using currently available digital signal processor devices. The Naval Command, Control and Ocean Surveillance Center Research, Development, Test and Evaluation Division developed a linear frequency modulated (LFM) radar for ISAR imaging of aircraft. Motion compensation and imaging techniques were developed to process data collected by this radar. These algorithms are categorized according to the accuracy of the motion estimate they provide. The performance of the algorithms is measure when they are applied to data collected using the NRaD LFM radar. The implications of these results for more general applications are examined.

In a target-rich battlefield environment, a shipboard or an airborne radar must maintain situational awareness while tracking and identifying targets. Often the opportunity to dwell on each target long enough for confident identification via high resolution SAR/ISAR imaging will not exist, especially for those engagement geometries where the relative translational motion of the aircraft does not result in large rotation rates. Inadvertent aircraft tactical dither often generates enough target rotational during a brief imaging interval to allow the formation of an ISAR image with low crossrange resolution. We have developed an automated identification procedure that utilizes this resolution, along with high range resolution, to produce confident target identification. The advanced signal processing algorithms employed extract feature measurements from the complex ISAR image. including accurate measurements of the two-dimensional positions, amplitudes and range extents of the dominant target scatterers. A deformable template matching procedure is used to correlate these 'measured features' with those predicted for each candidate aircraft in a database generated from readily available diagrams, photographs and CAD models. After obtaining the optimal fit between the measured and predicted features for each candidate aircraft, the 'most likely' candidate is selected using a conventional Bayes classifier.

Conventional radar processors use the Fourier transform to retrieve Doppler information. In order to use the Fourier transform adequately, some restrictions must be applied: the scatterers must remain in their range cells and their Doppler frequency contents should be stationary during the entire imaging time duration. If the Doppler frequency contents are time-varying, the Doppler spectrum obtained from the Fourier transform becomes smeared, and, thus the image resolution is degraded. However, the restrictions on the Fourier processing can be lifted if the Doppler information can be retrieved with the time-frequency transform which does not require a stationary Doppler spectrum. Therefore, the image blurring caused by the time- varying Doppler spectrum can be resolved without applying sophisticated motion compensation techniques to the individual scatterers.

In previous work, the authors have proposed the compensation of translational motion of an object via time-frequency filtering in a way that separation of scatterers can be done and compensation can be achieved selectively. In this paper, the effect of filtering in the time-frequency domain is studied focusing on the effects that this processing introduces to the phase of the scatterer response. The instantaneous frequency associated with one scatterer in the image is estimated and compared to the estimated frequency obtained for the case where only that scatterer is present by itself. Deviations caused by these two computations are presumed to affect the final compensation of an image regardless of which compensation technique is used.

An effective two-dimensional polar format algorithm based on the circular sampling theorem is implemented and tested. The algorithm interpolates samples from a polar to a rectangular raster for the purpose of focusing ISAR imager. The imagery are generated from samples collected in the frequency space utilizing a uniform polar set of coordinates. An example of an extended target is offered to show the versatility of the algorithm. In addition, a point target model is used to test its effectiveness. The distortion introduced by interpolation is calculated and compared to errors introduced by two standard interpolation techniques. Experimental data provided by the Pacific Missile Test Center was used to test the proposed algorithm.

In this article, we present a method for estimating the translational motion parameters of a target from its ISAR signature. The method exploits the phase of the target's frequency response and does not require 2D Fourier processing. The basis for this method is a phase difference indicator which converges to an absolute minimum when the phase is compensated with the true values of the motion parameters. The analysis of the phase difference indicator leads to an algorithm for motion parameter optimization. Processing of simulated and experimental ISAR signatures demonstrates that the phase difference method is extremely computationally efficient and equally accurate when compared to robust techniques based on entropy or Fisher information indicators.

The problem of image analysis when the image is complex- valued is becoming increasingly important in problems where the complex nature of the scattering mechanism cannot be ignored. Such problems include automatic target recognition and cross-section reduction. Traditional approaches have sought to apply standard methods in intensity-based imaging to these problems by discarding the inherent phase content. Other methods have been developed but are often ad hoc and insufficiently motivated. We will examine this problem and some of the recent approaches suggested for its solution.

3D images have been produced from SAR data limited to curved aperture paths by imaging techniques usually associated with two dimensional image resolution enhancement. CLEAN and IMP/RELAX algorithms have produced images for two different measurement scenarios. Dynamic range of CLEAN images are dependent upon level sat which artifacts occur although valid scatterers appear below such levels. Artifacts occur due to focusing errors in IMP/RELAX images and in addition due to sidelobe leakage effects in CLEAN images. These are accentuated by the very large sidelobes of shallow apertures. Autofocused images are significantly sharper than those from initial aperture estimates. Methods to expand the processed image volume will extend the general usefulness of this technique as will methods to estimate and display subtle variations in fully populated 3D image volumes. For these purposes resampling of data onto a rectilinear grid may allow use of image estimation techniques already proven useful for 2D imaging.

Using a 3D formulation of inverse synthetic aperture (ISAR), we show how the arbitrary formatting capability of a time- integrating opto-electronic processor can be used to format the data in such a way as to focus the image. The focusing parameters are found from prominent points in the radar data itself assuming a rigid body object. The opto-electronic processor is particularly suited for generated image data such as found with ISAR since it can arbitrarily place data points as needed depending on the object motion without need of interpolations. The processor is a time-integrating architecture that uses direct laser diode intensity modulation for signal insertion, acousto-optic scanners for arbitrary formatting, and a modified Kosters interferometer for stable 2D Fourier transformation. The processor also automatically generates multi-resolution imagery as a by- product of its generation of the final fine resolution image. Thus, there is a sequence of values available for each image pixel that are available for enhanced multi- resolution image processing. This research is funded by the Office of Naval Research.

The ImSyn^TM Processor is an optoelectronic signal processor developed by Essex Corporation to accelerate coherent imaging processes. This paper focuses on the application of the ImSyn Processor to SAR imaging where severe range differential curvature is present. This occurs in SAR systems imaging large scenes with fine resolution, foliage penetrating (FOPEN) radar and ground penetrating radar. Application of the range migration algorithm removes the differential range curvature but results in a non- uniform or warped frequency space. The ImSyn processor operates directly on the frequency data permitting a discrete Fourier transform in warped frequency space without data interpolation. Both the range migration algorithm and the standard polar formatting algorithm benefit from the increased speed and resolution available from the ImSyn processor. A discussion of the ImSyn processor, the range migration algorithm and an example of a FOPEN image processed on our prototype system are presented.

We previously reviewed the architecture and basic analytic results for a hybrid optical-digital processor capable of generating target range Doppler profiles in real time. Here we describe the optical design and present preliminary results for the optical correlator portion of the hybrid system. Phase control, fringe stability, and preliminary correlation data for the optical system are reported.

We have configured an adaptive multi-layer optical classifier as a real-time radar target recognition system to recognize isolated aircraft targets from training profiles with varying orientation and/or range from the radar. The resulting system demonstrates the successful application of a real-time adaptive optical computing system to a challenging temporal signal processing problem with time- bandwidth product requirements too demanding for alternative approaches. In this paper we describe the multi-layer classifier in detail and present classification results of using the optical system to learn form example time- frequency representations of aircraft radar range profiles.

This paper summarizes our studies of a time-integrating acousto-optic processor that offers a compact solution to detecting and analyzing wideband spread-spectrum signals for variety of radar, navigation or communication applications. The heart of the processor is a 2D quadrature radio- frequency time-integrating acousto-optic correlator complemented by a subsequent digital processor. Quadrature envelope processing allows the number of delay channels to be significantly reduced. The correlator can have a processing bandwidth up to 500 MHz, and the digital processor is used to perform nonlinear detection and get spectrum information. We discuss the general structure of the processor, possible operating modes and corresponding post-detector processing algorithms. Theoretical points include sensitivity, dynamic range and frequency resolution. Preliminary experimental results obtained in the UHF range are also presented.

We report on the first demonstration of a 2D, fiber-optic, true time-delay control system capable of squint-free steering of an ultrawideband array. The system is based on a novel fiber-optic prism technique with separate azimuth and elevation control stages. In this technique, high dispersion fiber is used to provide a wavelength-dependent time-delay between the rows and columns of the array. The fiber-optic control system was used to drive a 4 by 4 array of flared- notch elements. The system was tested in a compact radar range and 2D array patterns were obtained. The system demonstrated independent +/- 30 degrees elevation steering of a 4 by 4 array over a 6 to 18 GHz bandwidth with no observable squint. We believe this is the first demonstration of such a system. The design of the demonstrated system allows for transitioning to larger, real-world ultrawideband array transmitters.

Emerging missions for shipboard defense and tactical ballistic missile defense require the support of wideband, multi-function radars capable of concurrently performing hemispherical surveillance, tracking and simultaneous illumination of multiple targets. Active phased array antennas to support these missions are limited by their cost, bandwidth and aperture weight. As mission requirements become more demanding the integration of photonics into phased arrays, which promises increased bandwidth, decreased aperture weight and less complex transmit/receive modules, must be pursued. As part of the Office of Naval Research's Accelerated Capabilities Initiative Raytheon, supported by the University of Connecticut, is developing a novel photonic antenna architecture for the control of active phased arrays. The photonic architecture is optically non- coherent and achieves a reduction in hardware complexity, and therefore array cost, via device sharing which is facilitated through wavelength division multiplexing (WDM). By properly configuring the photonic architecture, WDM represents a beneficial compromise between hardware complexity and array performance. To realize the photonic architecture the University of Connecticut is developing novel multiple quantum well photonic devices, including electronically tunable lasers and filters and a broadband amplitude modulator. This paper will discuss the antenna architecture wavelength division multiplexing and the enabling photonic devices.

A 3-bit binary photonic delay line is demonstrated at 1 GHz using a directly modulated semiconductor laser and remote interconnection fiber-optics. Three types of freespace delay bit geometries are tested for 5.686 ns, 1.667 ns, and 8.83 ps delay bits. This photonic delay line uses ferroelectric liquid crystal optical on/off devices for optical path switching and active polarization noise filtering. 3D imaging optics and antireflection coated optics are successfully used to minimize photonic delay line insertion losses and interchannel crosstalk. The 3-bit system is fully characterized for measured and designed performance.

We present new optoelectronic architectures, based on parallel delay lines, performing programmable filtering of microwave signals. According to current performances of optoelectronic components, they can process optically carried microwave signals over frequency bandwidths as large as 20 GHz, with a time-frequency product up to 10³. The operating principle of these structures is detailed and followed by the preliminary experimental demonstration at 1.3 GHz of a 53 dB rejection filter.

We present a novel and efficient approach to true-time-delay (TTD) beamforming for large adaptive phased arrays with N elements, for application in radar, sonar, and communication. This broadband and efficient adaptive method for time-delay array processing algorithm decreases the number of tapped delay lines required for N-element arrays form N to only 2, producing an enormous savings in optical hardware, especially for large arrays. This new adaptive system provides the full NM degrees of freedom of a conventional N element time delay beamformer with M taps, each, enabling it to fully and optimally adapt to an arbitrary complex spatio-temporal signal environment that can contain broadband signals, noise, and narrowband and broadband jammers, all of which can arrive from arbitrary angles onto an arbitrarily shaped array. The photonic implementation of this algorithm uses index gratings produce in the volume of photorefractive crystals as the adaptive weights in a TTD beamforming network, 1 or 2 acousto-optic devices for signal injection, and 1 or 2 time-delay-and- integrate detectors for signal extraction. This approach achieves significant reduction in hardware complexity when compared to systems employing discrete RF hardware for the weights or when compared to alternative optical systems that typically use N channel acousto-optic deflectors.

We present on the use of coherent optical beam combining and signal processing in a fiber-optic phased-array-antenna systems. Mutually coherent links lose phase registration after the signals propagate through the fiber path lengths subject to thermal and stress-induced variations. These path length changes can be compensated with phase corrections to the optical carrier in tan AM or PM modulated signal. In a true time delay fiber optical feed subsystem, the phase corrected outputs can be coherently combined to allow a lens to collect almost all the main beam energy into a single mode output, leading to substantially improved noise figure and dynamic range.

A high bandwidth, large degree-of-freedom photorefractive phased-array antenna beam-forming processor which uses 3D dynamic volume holograms in photorefractive crystals to time integrate the adaptive weights to perform beam steering and jammer-cancellation signal-processing tasks is described. The processor calculates the angle-of-arrival of a desired signal of interest and steers the antenna pattern in the direction of this desired signal by forming a dynamic holographic grating proportional to the correlation between the incoming signal of interest from the antenna array and the temporal waveform of the desired signal. Experimental results of main-beam formation and measured array-functions are presented in holographic index grating and the resulting processor output.

This paper presents an analytical derivation of the effects of Doppler shift on the optimal signal processing of frequency-stepped CW radar data. It is shown that successive CW measurements produce a shift in the global minimum of the objective function which can be tracked as a function of time. The amount of time shift in the global minimum can be used to extract the target velocity. Once the target's velocity has been established, accurate estimates of target position are obtained. Simulation results are presented that illustrate the process of velocity and range information extraction from the Doppler-shifted frequency-stepped CW data.

High range resolution capability provides many advantages in radar but it requires large bandwidth systems. However with step frequency radars it is possible to achieve high resolution of wideband systems with many of the advantages of narrow band systems. In this paper ambiguity function is employed to study step frequency radar. The mathematical expression is derived for the ambiguity function of step frequency radar receiver and processor. The processor functions included are range gating, weighting, fill pulses and inverse discrete Fourier transform. Ambiguity diagrams are plotted for step frequency receiver processor and comparison is made with traditional constant frequency radar.

With reference to the air target detection of ultra- wideband(UWB)/impulse radar, we discussed transient signal processing techniques. In weak UWB signal detection, wavelet transforms and high order spectrum estimation techniques were preferred. In target characteristic analysis, two algorithms of impulse response deconvolution, MCGM and DPREM, were presented. In this paper, a time domain bispectrum estimation algorithm was used to analyze target impulse response, which could estimate accurately local scattering distribution of complex target. A free field IR experimental system was used which was laid out in an anechoic chamber. With this system, we measured the response of several target models and a scale aircraft.

On the topic of the passive position location and tracking, some people have published many papers. In this paper, we will discuss the passive position location and tracking based on the extended Kalman filter. In our experiments, the watching stations are stationary, but the targets are moving. We make the two watching stations work together at first, and obtain the original values for the extended Kalman filter. When the estimating or filtering error of Kalman filter is small enough, then we can make tow stations work separately. So each station can at least track one target. In the end of this paper, we give out some simulation experiment results, from which we can verify the method we proposed.

Exact and approximate methods are developed to determine radar antenna polarizations that maximize power contrast between two scatterers. Antenna polarizations to maximize received power are also obtained. These polarizations will maximize target-noise ratio and the ratio of target power to power from unpolarized clutter. Matrices that represent unpolarized clutter are given. It is shown that a backscattering covariance matrix to represent unpolarized clutter for any incident wave does not exist. Target matrices used are the bistatic and backscattering covariance matrices, the Kennaugh matrix, and the Graves power scattering matrix.

A compact and affordable photonic true-time-delay (TTD) beam steering device for phased array antenna applications using multiplexed substrate-guided wave propagation is presented. The TTD design uses holographic input and output couplers to change the direction of beam propagation as well as optical fanout. Optical delays of various delay lines can be adjusted easily through the substrate thickness and the total internal reflection angle inside the substrate material. Broadband microwave signals for feeding the radiating elements are generated through optical heterodyne technique and they are detected by metal-semiconductor-metal detector arrays. The physical aspects of phase-shifters and true-time-delays are first introduced. Then design issues on the photonic TTD architecture and practical constraints on making holographic grating couplers are discussed, especially concerning with recording gratings on DuPont photopolymer materials. Finally, the generation and detection of high frequency microwave signals up to 25 GHz by optical heterodyne techniques are illustrated.

A novel technique has been developed to inject coherence in an ultrawideband radar system that transmits white Gaussian random noise. Coherence is introduced in the system by performing heterodyne correlation of the received signal with the time-delayed replica of the transmit signal. This operation preserves the phase of the reflected signal which is lost in a traditional homodyne correlation receiver. Knowledge of the phase of the received signal permits the configuration of the system as a spaced antenna interferometer for azimuthal scanning and transversal speed estimation. This paper describes the basic theory of random noise radar interferometry, and presents first results obtained using the University of Nebraska's 1-2 GHz random noise radar system configured as a radar interferometer.