Search

For small radar platforms, the increase in signal-to-interference-plus-noise ratio (SINR) needed to support effective operation must come through integration of target signal energy collected over a long dwell time. Conventional radar processing assumes a linear-phase signal model and utilizes Fourier-based methods to coherently integrate signal energy. Over an extended dwell, the target signal generally includes multiple nonlinear-phase components which limit the effectiveness of conventional methods. An algorithm is presented that estimates the linear- and nonlinear-phase components of the extended-dwell-time target signal in a multistage process. The combined phase components form the signal model used in a filter that achieves near optimal matching performance for extended dwell times over a wide range of target parameters. Results are presented for both noise-limited and clutter-limited environments. For the specific conditions discussed, typical increase in output SINR for a 500 ms dwell time is 12 dB over the conventional coherent processing methods.

Improvements in target detection are typically achieved by increasing the signal-to-interference-plus-noise ratio (SINR) of the received signal. For a small radar platform with power and aperture constraints, increases in SINR must come through temporal integration. Two techniques are presented here for enhancing the integration of temporal signal energy to improve detection of weak, slow-moving targets over conventional methods. Both techniques involve modifications to pre-Doppler space-time adaptive processing (STAP), a reduced-dimension STAP method that is implementable in real-time on a power-constrained platform. The two modifications include (i) development of an alternative pre-Doppler temporal weighting method, based on linear prediction, to increase the SINR of the pre-Doppler temporal output signal and (ii) development of an extended dwell temporal processing (EDTP) algorithm to integrate temporal signal energy collected over a long dwell. The EDTP algorithm is based on frequency domain analysis of the extended dwell time signal; the detection decision exploits the differences in the distribution of signal energy and noise energy over an extended dwell. EDTP algorithm performance is superior to traditional temporal processing methods in many practical scenarios: the EDTP algorithm detection rate is three times the detection rate of non-coherent integration for very low-velocity, low-SNR targets.

The authors derive a measure suitable for identifying sensor placements that yield high target classification rates in a distributed radar environment. A distance measure is derived as an approximation lower bound of the Kullback–Leibler (KL) divergence between two probable target classes. A relationship between the KL divergence approximation and probability of correct classification (PCC) is demonstrated and used to identify sensor placements likely to yield higher PCC values. Two algorithms are proposed, optimal and approximate, for identifying good placements of a single transmitter and two receivers. A physics-based scattering model is employed to generate reflectivity data suitable for multi-sensor, multi-static classification analysis. KL and PCC data sets are analysed to demonstrate the effectiveness of the proposed distance measure and algorithms.

Reducing the size, weight, and power (SWAP) of airborne radar systems improves efficiency and decreases operational cost. Platform SWAP requirements are directly related to the complexity of algorithms used to process the received signal. Several reduced dimension space-time adaptive processing (RD-STAP) algorithms, which significantly decrease computations over traditional STAP, have been shown to minimally impact GMTI performance in homogeneous clutter. In both heterogeneous clutter and dense target environments, training with data whose characteristics match those of the cell under test may be required to maintain acceptable performance; however, selecting appropriate training data requires additional computations that increase SWAP. This paper identifies and quantifies trades between performance and computations for several RD-STAP algorithms and training methods in challenging signal environments.

In this paper, we use an information theoretical sensor placement algorithm to assess the impact of target camouflage, concealment, and deception (CCD) effects on classifier performance. Physics-based target models are constructed to exhibit varying CCD effects of a single target class. An information theoretical sensor placement algorithm is used to identify potential sensor locations yielding highly probable discrimination of test targets representing non-CCD and CCD targets. Platforms are positioned according to the identified sensor locations for classification processing. Classification performance results are presented and discussed in the context of the modeled CCD effect. Results demonstrate the effectiveness of the placement algorithm to identify sensor locations void of the intended CCD effects.

Output signal-to-interference-plus-noise ratio (SINR) can be increased by extending the dwell time or coherent processing interval (CPI) over which target signal energy is integrated. However, over a long dwell time, nonlinear phase components in the slow-time signal limit the effectiveness of traditional Fourier-based processing methods to coherently integrate signal energy; thus, signal energy is often coherently integrated on a sub-CPI basis. This paper investigates a new extended dwell temporal signal processing algorithm that utilizes the signal energy in adjacent sub-CPIs in a unique way to improve the detection of weak, slow-moving targets. In many cases, the extended dwell algorithm more than doubles the probability of detection of low radial velocity targets over conventional noncoherent integration.

Temporal weights used in element-space pre- Doppler STAP ideally maximize the magnitude and frequency coverage of the pre-Doppler temporal output signal. Maximizing the magnitude of the temporal output signal is critical for detection of weak targets, while broad frequency coverage is essential for detection of low radial velocity targets. Traditional pre-Doppler STAP uses binomial filter coefficients, which act as a high-pass filter, for the temporal weights. A comparison of the traditional weighting method and an alternative method, which uses linear prediction to determine the temporal weights, shows that weights determined with linear prediction often contribute to a higher output SINR than do binomial temporal weights. In addition, linear prediction based temporal weights are determined adaptively from the data which offers a significant advantage in flexibility over the traditional method whose weights are based on a specific collection geometry.

In this paper, we use a probabilistic divergence measure to identify radar sensor placements that yield high target classification rates. The derived divergence measure uses a lower bound of the Kullback-Leibler divergence to recognize significant differences in aspect-dependent target class probability distributions. Monte Carlo simulations are performed at various noise levels to demonstrate the similarity between the divergence measure and probabilities of correct classification (PCC). High range resolution (HRR) profiles are used as inputs to a multi-sensor classifier to identify the most probable target classification. Synthetic targets with dominant scatterers are employed to show the benefits of exploiting spatial diversity from prominent target features.

Recently, significant progress has been made in the development of physics-based, knowledge-aided (KA) signal processing strategies supported by improvements in real-time embedded computing architectures. These developments provide designers of advanced sensor systems an unprecedented degree of flexibility when implementing next generation adaptive sensor systems. In the case of radar, this has been manifested in the first ever, real-time, KA space-time adaptive processing (KA-STAP) system for advanced clutter/interference suppression. This paper provides exemplars of real-world effects giving rise to the need for "intelligent" adaptation schemes and overviews the KA approach to sensor signal processing in some detail. Moreover, we survey a collection of papers describing recent KA sensor research that follow in this issue

This paper introduces a parametric covariance estimation scheme for use with space-time adaptive processing (STAP) methods operating in heterogeneous clutter environments. The approach blends both a priori knowledge and data observations within a parameterized model to capture instantaneous characteristics of the cell under test (CUT) and reduce covariance errors leading to detection performance loss. We justify this method using both measured and synthetic data. Performance potential for the specific operating conditions examined herein include: 1) averaged behavior within roughly 2 dB of the optimal filter, 2) 1 dB improvement in exceedance characteristic relative to the optimal filter, highlighting improved instantaneous capability, and 3) impervious ness to corruptive target-like signals in the secondary data (no additional signal-to-interference-plus-noise ratio (SINK) loss, compared with 10 dB or greater loss for the standard STAP implementation), with corresponding detections comparable to the optimal filter case

This tutorial provides a brief overview of space-time adaptive processing (STAP) for radar applications. We discuss space-time signal diversity and various forms of the adaptive processor, including reduced-dimension and reduced-rank STAP approaches. Additionally, we describe the space-time properties of ground clutter and noise-jamming, as well as essential STAP performance metrics. We conclude this tutorial with an overview of some current STAP topics: space-based radar, bistatic STAP, knowledge-aided STAP, multi-channel synthetic aperture radar and non-sidelooking array configurations.

Space-time adaptive processing (STAP) is an important airborne radar technique used to improve target detection in clutter-limited environments. Effective STAP implementation is dependent on accurate space-time covariance matrix estimation. Heterogeneous clutter, including spiky, spatial clutter variation, violates underlying STAP training assumptions and can significantly degrade corresponding detection performance. This paper develops a spiky, space-time clutter model based on the K-distribution, assesses the resulting impact on STAP performance using traditional methods, and then proposes and evaluates the utility of the knowledge-aided parametric covariance matrix estimation (KAPE) method, a model-based scheme that rapidly converges to better represent spatial variation in clutter properties. Via numerical simulation of an airborne radar scenario operating in a spiky clutter environment, we find substantial improvement in probability of detection ( P D ) for a fixed probability of false alarm ( P F A ) for the KAPE method. For example, in the spiky clutter environment considered herein, results indicate a P D of 32% for traditional STAP and in excess of 90% for KAPE at a P F A of 1E-4, with a corresponding difference of 11.5 dB in threshold observed from exceedance analysis. The proposed K-distributed spiky clutter model, and application and assessment of KAPE as an ameliorating STAP technique, contribute to an improved understanding of radar detection in complex clutter environments.

This paper considers the problem of screeningkmultivariate normal populations (secondary data) with respect to a control population (primary data) in terms of covariance structure. A screening procedure, developed based upon statistical ranking and selection theory, is designed to include in the selected subset those populations which have the same (or similar) covariance structure as the control population, and exclude those populations which differ significantly. Formulas for computing the probability of a correct selection and the least favorable configuration are developed. The sample size required to achieve a specific probability requirement is also developed, with results presented in tabular form. This secondary data selection procedure is illustrated via an example with applications to radar signal processing.

This paper describes an innovative concept for knowledge-based control of space-time adaptive processing (STAP) for airborne early warning radar. The knowledge-based approach holds potential for significant performance improvements over classical STAP processing in nonhomogeneous environments by taking advantage of a priori knowledge. Under this approach, knowledge-based control is used to direct pre-adaptive filtering, and to carefully select STAP algorithms, parameters, and secondary data cells.

Systems Engineering Research Center has developed a novel cyber security concept for embedding security solutions into systems called System-Aware cyber security. The goal of the System-Aware program is to develop low cost methods of protection against cyber exploits by our adversaries. To demonstrate the effectiveness of the System-Aware design patterns, specific ones were developed for an unmanned aerial vehicle (UAV) application. The application to UAV-based systems was inspired by the wide variety of subsystems that are used in UAV configurations, the range of potential cyber-attacks that can seriously impact the critical missions of these systems, the significant power, space and performance constraints that System-Aware designs must address in order to operate in UAV-based configurations. This report is the phase-II of the Security Engineering project where conducting a flight demonstration of the System-Aware Sentinel was planned. It consists of activities necessary to integrate the results of the Phase-I effort into the GTRI Aerial Unmanned Sensor System (GAUSS) aircraft in order to create a flight-ready demonstration. The GAUSS platform is a small research UAV with a commercial off-the-shelf autopilot system and camera gimbal. The demonstration will show how the System-Aware approach can be used to thwart cyber-attacks against autopilot and sensor systems.

Surface moving target indication (SMTI) involves searching Earth's surface for moving objects using a dedicated radar mode. Ground-moving target indication (GMTI) is a subordinate, commonly referenced mode implying the detection, location, and discrimination of vehicles and personnel (dismounts) against rural, suburban, and urban land settings. The general class of SMTI radar modes also includes searching for vessels against sea, lake, and riverine backgrounds. An SMTI platform can be an aircraft, an unmanned aerial system (UAS), a satellite, an aerostat, or a tower.

Space-time adaptive processing (STAP) is an important radar technology. It is a cornerstone in the design of modern moving target indication and imaging radar systems. Specifically, STAP is a multidimensional filtering technique that mitigates the influence of clutter or radio frequency interference on principal radar products, viz. radar detections or images. This work serves as a tutorial reference of fundamental STAP concepts and techniques. The interested reader will encounter the basic STAP formulation, common STAP performance metrics, signal models, interference mitigation approaches, an application example corresponding to Doppler spread clutter mitigation in airborne radar systems, current challenges, and some implementation issues. The fundamentals presented herein are extensible to a range of radar signal processing problems.

This paper examines the detection of weak moving targets using long-dwell, multi-channel radar. The impact of unknown target motion on detection performance is addressed. Signal-to-noise ratio (SNR) loss caused by mismatch between the matched filter and the actual target phase history - as might occur when using conventional space-time adaptive processing (STAP) - is quantified, and major factors contributing to this loss are identified. Limits on dwell time and target motion are established for conventional STAP. Results in this paper form a basis for long-dwell algorithm development.

Radar target classification is examined from the viewpoint of improving classification performance through the use of spatial diversity. Improved radar target classification has been demonstrated previously by using at least one additional perspective in a generic environment but the impact of sensor placement has been less studied. In this paper, we examine the use of multiple high range resolution (HRR) profiles to demonstrate how selection of sensor locations can improve classification rates. Specifically, performance improvements are demonstrated after identifying the optimal set of perspectives and employing a simple decision fusion network (DFN) algorithm for defined signal-to-noise (SNR) levels. We show percentages of correct classification (PCC) can be maintained in scenarios where SNR has been reduced by up to 9 dB on a single sensor basis.

The Systems Engineering Research Center (SERC) has developed a novel cybersecurity concept for embedding security solutions into systems called System-Aware Cybersecurity. The overall goal of the System-Aware program is to develop low cost methods of protection against cyber exploits by our adversaries. Development of a prototype security system for securely monitoring an autonomous surveillance system on board an unmanned aerial vehicle for possible cyber attacks using reconfiguring systems for Sentinel-based architecture. Exploring decision support methodologies for determining on a mission basis the most critical system functions to secure employing attack tree tools and SysML/UML tools. Developing cyber security CONOPS for operation of UAV's that are possibly under attack. Exploring the opportunity to apply private Cloud capabilities as a Sentinel for monitoring ground-based systems so as be able to readily employ moving target and diversity solutions to secure the Sentinel and to monitor Cloud performance as a means for detecting possible cyber attacks. The major deliverables of the project were prototype-based and simulation experiments that result in enhanced understanding of requirements for cyber defense of systems and design patterns that can be reused across multiple system types.

This paper describes the results of a study to evaluate ground moving target indication (GMTI) challenges in mountainous terrain and identify new exploitation methods to enhance the capabilities of a nominal X-band radar system. Through simulation and analysis we describe the difficulties inherent in space-time adaptive processing (STAP) GMTI in highly mountainous terrain. We use intermediate performance metrics, such as detection rate and signal-to-interference-plus-noise ratio (SINR) losses over range and Doppler, to identify physical mechanisms leading to STAP inefficiency and characterize the benefits of improvement strategies, such as Power Comparable Training (PCT), a data-dependent training method. The focus of this effort is to understand factors adversely affecting the detection of surface targets, including vehicles and dismounts; document the performance impacts; and, identify a plausible solution space.

Humans are difficult targets to detect because they have small radar cross sections (RCS) and move at low velocities. Consequently, they are masked by Doppler spread ground clutter generated by the radar bearing platform motion. Furthermore, conventional radar-based human detection systems employ some type of linear-phase matched filtering, whereas most human targets generate a highly nonlinear phase history. This work proposes an enhanced, optimized, nonlinear phase (EnONLP) matched filter that exploits knowledge of human gait to improve the radar detection performance of human targets. A parametric model of the expected human response is derived for multi-channel radar systems and used to generate a dictionary of human returns for a range of possible parameter variations. The best linear combination of projections in this dictionary is computed via orthogonal matching pursuit (OMP) to detect and extract features for multiple targets. Performance of the proposed EnONLP method is compared with that of traditional space-time adaptive processing (STAP) and a previously derived parameter estimation-based ONLP detector. Results show that EnONLP exhibits a detection probability of about 0.8 for a clutter-to-noise (CNR) ratio of 20 dB and input signal-to-noise ratio (SNR) of 0 dB, while ONLP yields a 0.3 and STAP yields a 0.18 probability of detection for the same false alarm rate.

Modern radar systems are highly complex, leveraging the latest advances in technology and relying on sophisticated algorithms and processing techniques to yield exceptional products. Principals of Modern Radar is the first in a series, covering basic radar concepts, radar signal characteristics, radar subsystems, and basic radar signal processing. This text is the second in the series and contains advanced techniques, including the most recent developments in the radar community. Specifically, much of Principles of Modern Radar: Advanced Techniques discusses radar signal processing methods essential to the success of current and future radar systems. Applying these techniques may require specific hardware configurations or radar topologies, as discussed herein.

Aerospace radar systems must detect targets competing with strong clutter and jamming signals. For this reason, the radar system designer incorporates a mechanism to suppress such interference. A familiarity with the application of space-time degrees of freedom - shown herein to greatly enhance detection in interference-limited environments- is thus essential. This chapter discusses the important role of space-time adaptive processing in moving target indication radar.

Many current radar-based human detection systems employ some type of Doppler or Fourier-based processing, followed by spectrogram and gait analysis to classify detected targets. However, Fourier-based techniques inherently assume a linear variation in target phase over the aperture, whereas human targets have a highly nonlinear phase history. This mismatch leads to significant loss in SNR and integration gain. In this paper, two novel human-modeling based non-linear phase detectors are presented. The first (ONLP) computes maximum likelihood estimates of unknown parameters of a model of the human torso response, while the second (EnONLP) stores the expected returns of a 12-point model for each combination of model parameter values in a dictionary and uses orthogonal matching pursuit to find the optimal sparse approximation to the data. The performance of ONLP, EnONLP, and conventional STAP is compared and application to target characterization discussed.

Many current radar-based human detection systems employ some type of Doppler or Fourier-based processing, followed by spectrogram and gait analysis to classify detected targets. However, in Fourier-based techniques the maximum output signal-to-noise ratio (SNR) is given by targets whose target phase is linear. On the contrary, the phase variation of the human target response is nonlinear. This difference causes a significant loss in SNR, and therefore detection performance. In this paper, two novel, nonlinear phase detector designs based on human modeling are presented. In the first method, only the human torso reflections are modeled and unknown model parameters computed using Maximum Likelihood Estimation. In the second method, the entire human body is modeled as a different parametric model. The expected radar response for each combination of parameter values is stored in a database. An optimal sparse approximation to the data is found using Orthogonal Matching Pursuit. The performance of the proposed techniques and optimal space-time adaptive processing algorithm is compared and target characterization applications examined.

Many current radar-based human detection systems employ some type of Doppler or Fourier-based processing, followed by spectrogram and gait analysis to classify detected targets. However, Fourier-based techniques inherently assume a linear variation in target phase over the aperture, whereas human targets have a highly nonlinear phase history. This mismatch leads to significant loss in SNR and integration gain. In this paper, an Enhanced Optimized Non-Linear Phase (EnONLP) detector is proposed that employs a dictionary to store possible target returns generated from the human model for each combination of parameter values. An orthogonal matching pursuit algorithm is used to compute a sparse approximation to the radar return that is optimal in the least squares sense. Performance of the EnONLP algorithm is compared to that of a parameter-estimation based algorithm and conventional, fully-adaptive STAP.

We investigate the impact of heterogeneous clutter on STAP performance. The K-distribution is a good fit for many clutter scenes of interest and will be the main focus of this paper. We introduce a cell-based model that can simulate realistic spikiness of heterogeneous clutter, allowing us to empirically measure the losses in detection performance. In particular, it is shown that, for a fixed false-alarm probability of 10−4, the adaptive matched filter attains a probability of detection of 0.55 in severely heterogeneous clutter, compared to the achievable probability of detection of 0.9 in homogeneous clutter. We subsequently investigate the causes for this deterioration in performance.

Future advanced radar systems must detect targets of diminishing radar cross section (RCS) at low radial velocity, in demanding clutter and interference environments. Presently, a deficiency in radar detection performance exists between the capabilities of synthetic aperture radar (SAR) for fixed target indication and space-time adaptive processing (STAP) for ground moving target indication (GMTI) of targets with low ground track velocity. Dismounts, individuals or groups running, walking, or crawling, constitute a class of targets that falls into this netherworld between SAR and STAP. While possessing low RCS levels and radial velocities, dismount detection is rendered even more challenging due to their complicated non-linear phase histories that give rise to significant micro-Doppler energies. In this paper we develop a physiological human-gait model for multi-channel moving radar platforms. We characterize the dismount detection performance of a notational UAV system using linear phase, quadratic phase and sinusoidal phase filters. Finally, we summarize our results and present areas of future work.

Traditionally, moving target indication (MTI) modes tailored to wide-area surveillance (WAS) have employed waveforms with bandwidths matched to the range extent of targets of interest. Matching range resolution to expected vehicle lengths serves to minimize interfering clutter returns while preserving available target power, thereby maximizing target probability-of-detection. Recent advances in waveform generator technology and high-speed digital signal processors for matched filtering have encouraged radar designers to consider wideband waveforms for WAS MTI needs. While providing information for identification and imaging, wideband waveforms complicate the detection process, as an extended target will be over-resolved; that is, returns from a vehicle will be distributed over multiple, fine-resolution range bins. Algorithms for combining these signal contributions are required for detection and angle estimation. Given such algorithms, one wonders whether over-resolving targets leads to detection and angle estimation performance losses as compared to narrowband operation. In this paper we evaluate methods for collecting energy from an over-resolved target for detection and angle estimation processing. We bound the performance of these techniques on typical ground targets and show that the probability-of-detection and angle accuracy is improved through the use of wideband waveforms at high signal-to-noise levels.

In this paper we discuss a number of recent, important developments in adaptive radar. Specifically, we comprehensively identify critical trends and describe in some detail a number of corresponding solutions helping transition space-time adaptive processing (STAP) from theory to practice. Specific issues considered herein include: (1) the impact of clutter heterogeneity on detection performance; (2) the role of sensor deployment and the inducement of clutter nonstationarity; (3) the influence of aperture on system performance; and, (4) limitations imposed by computational burden on the deployable solution space. The intent of the paper is to juxtapose STAP theory and practice, with an emphasis on some developments of the past dozen years or so.

Conformal arrays possess certain desirable characteristics for deployment on unmanned aerial vehicles and other payload-limited platforms: aerodynamic design, minimal payload weight, increased field of view, and ease of integration with diverse sensor functions. However, the conformal arraypsilas nonplanar geometry causes high adaptive losses in conventional space-time adaptive processing (STAP) algorithms. In this paper, we develop a conformal array signal model and apply it to evaluate the performance of conventional STAP algorithms on simulated ground clutter data. We find that array-induced clutter nonstationarity leads to high adaptive losses, which greatly burden detection performance. To improve adaptive performance, we investigate the application deterministic and adaptive angle-Doppler warping techniques, which align nonstationary clutter returns. Through the application of these techniques, we are able to nearly fully mitigate the nonstationary behavior yielding performance similar to that of a conventional planar array.

Radar offers unique advantages over other sensors in the human detection problem, such as remote operation during virtually all weather and lighting conditions. Many radar-based human detection systems today employ Fourier analysis, such as spectrograms. However, spectrograms perform poorly in high clutter environments. Also, an inherent SNR loss is caused by the implicit assumption of linear target phase. In this paper, human modeling is used to derive a more accurate non-linear approximation to the true non-linear target phase and the likelihood ratio is optimized over unknown parameters to enhance detection performance. Performance is compared both analytically and through MATLAB simulations.