Article

Adaptive detection of range-spread target in compound-Gaussian clutter without secondary data

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

In this paper, we address the problem of detecting a range-spread target embedded in a non-Gaussian clutter with unknown covariance matrix and without using secondary data. We propose a new autoregressive method based on the generalised likelihood ratio test (GLRT) that requires only the cells under test. This method is used to derive two new detectors, corresponding to two different scenarios: a) when all range cells contain the target and share the same covariance matrix (homogeneous clutter), b) when different covariance matrices for different range cells are assumed (heterogeneous clutter). The proposed method is shown to outperform the state of the art on various scenarios in terms of false alarm probability and detection probability, especially in critical scenario as small data records or low number of secondary data. Finally, it exhibits the desired constant false alarm rate (CFAR) property.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... In [205], the nonstationary PAMF (NS-PAMF) and nonstationary normalized PAMF (NS-PAMF) were proposed for adaptive signal detection in hyperspectral imaging. There are many other parametric detectors, e.g., [206][207][208][209][210][211][212][213][214][215][216][217][218][219][220][221]. ...
Preprint
Full-text available
Multichannel adaptive signal detection jointly uses the test and training data to form an adaptive detector, and then make a decision on whether a target exists or not. Remarkably, the resulting adaptive detectors usually possess the constant false alarm rate (CFAR) properties, and hence no additional CFAR processing is needed. Filtering is not needed as a processing procedure either, since the function of filtering is embedded in the adaptive detector. Moreover, adaptive detection usually exhibits better detection performance than the filtering-then-CFAR detection technique. Multichannel adaptive signal detection has been more than 30 years since the first multichannel adaptive detector was proposed by Kelly in 1986. However, there are fewer overview articles on this topic. In this paper we give a tutorial overview of multichannel adaptive signal detection, with emphasis on Gaussian background. We present the main deign criteria for adaptive detectors, investigate the relationship between adaptive detection and filtering-then-CFAR detection, relationship between adaptive detectors and adaptive filters, summarize typical adaptive detectors, show numerical examples, give comprehensive literature review, and discuss some possible further research tracks.
... In [207], the nonstationary PAMF (NS-PAMF) and nonstationary normalized PAMF (NS-PAMF) were proposed for adaptive signal detection in hyperspectral imaging. There are many other parametric detectors, e.g., [208][209][210][211][212][213][214][215][216][217][218][219][220][221][222][223]. ...
Article
Full-text available
Multichannel adaptive signal detection jointly uses the test and training data to form an adap-tive detector, and then make a decision on whether a target exists or not. Remarkably, the resulting adaptive detectors usually possess the constant false alarm rate (CFAR) properties, and hence no additional CFAR processing is needed. Filtering is not needed as a processing procedure either, since the function of filtering is embedded in the adaptive detector. Moreover, adaptive detection usually exhibits better detection performance than the filtering-then-CFAR detection technique. Multichannel adaptive signal detection has been more than 30 years since the first multichannel adaptive detector was proposed by Kelly in 1986. However, there are fewer overview articles on this topic. In this paper we give a tutorial overview of multichannel adaptive signal detection, with emphasis on Gaussian background. We present the main deign criteria for adaptive detectors, investigate the relationship between adaptive detection and filtering-then-CFAR detection , relationship between adaptive detectors and adaptive filters, summarize typical adaptive detectors, show numerical examples, give comprehensive literature review, and discuss some possible further research tracks.
Article
We derive several parametric detectors for the case of a ground-based distributed multiple-input multiple-output (MIMO) radar system with co-located transmitters and receivers. The proposed detectors are based on exact and approximate generalized likelihood ratio tests. For the transmit-receive paths which start and end at the same node, we incorporate the constraint of spectral symmetry, meaning that the clutter returns are assumed to have a power spectral density that is symmetric around the zero Doppler, which is motivated by statistical analyses performed on measured clutter radar data. Our numerical simulations suggest that the proposed parametric detectors, which incorporate the physics-based spectral symmetry constraint on clutter returns, achieve better performance than previously investigated parametric detection approaches which do not incorporate this constraint.
Article
Adaptive detection of multichannel signals in Gaussian background is studied in this paper, for the case where the existence of training data is not assumed. Four new detectors are designed for this detection problem, by using the generalized likelihood ratio test, Rao test, Wald test and a reduced-dimension (RD) approach; their probabilities of false alarm (PFAs) and detection (PDs) are analytically deduced. These PFAs indicate that the four new detectors possess the constant false alarm rate properties against the noise covariance matrix. Experimental results show that the RD-based detector achieves larger (smaller) PDs than the other three new detectors if limited (sufficient) test data are available. When mismatched signals are encountered, the RD- and Rao-based detectors perform more robust and more sensitive, respectively, than the other two new detectors.
Thesis
In this thesis, we consider first the problem of low dimensional signal subspace estimation in a Bayesian context. We focus on compound Gaussian signals embedded in white Gaussian noise, which is a realistic modeling for various array processing applications. Following the Bayesian framework, we derive algorithms to compute both the maximum a posteriori and the so-called minimum mean square distance estimator, which minimizes the average natural distance between the true range space of interest and its estimate. Such approaches have shown their interests for signal subspace estimation in the small sample support and/or low signal to noise ratio contexts. As a byproduct, we also introduce a generalized version of the complex Bingham Langevin distribution in order to model the prior on the subspace orthonormal basis. Numerical simulations illustrate the performance of the proposed algorithms. Then, a practical example of Bayesian prior design is presented for the purpose of radar detection.Second, we aim to test common properties between low rank structured covariance matrices.Indeed, this hypothesis testing has been shown to be a relevant approach for change and/oranomaly detection in synthetic aperture radar images. While the term similarity usually refersto equality or proportionality, we explore the testing of shared properties in the structure oflow rank plus identity covariance matrices, which are appropriate for radar processing. Specifically,we derive generalized likelihood ratio tests to infer i) on the equality/proportionality ofthe low rank signal component of covariance matrices, and ii) on the equality of the signalsubspace component of covariance matrices. The formulation of the second test involves nontrivialoptimization problems for which we tailor ecient Majorization-Minimization algorithms.Eventually, the proposed detection methods enjoy interesting properties, that are illustrated on simulations and on an application to real data for change detection.
Article
Full-text available
The successful optimisation of the detection performance of maritime surveillance radars requires a detailed knowledge and understanding of both forward and backscattering from the ocean surface. Such an understanding enables the development of suitable signal processing techniques. In part I the authors discuss radar scattering from an ocean environment. It is shown, through the analysis of full-scale measurements, how the amplitude and correlation properties of high-resolution radar backscatter (sea clutter) can be accurately represented by the compound K-distribution model which has the unique characteristic of providing realistic performance predictions for a wide range of signal processing techniques. In addition to modelling noncoherent clutter, it is shown how the spectral and polarisation characteristics of coherently detected sea clutter relate to those observed in the noncoherent case. Results are also presented of forward-scattering and multipath propagation. This includes a consideration of the importance of the spatial and temporal coherence of the forward-scattered wavefront
Article
Full-text available
The authors consider the problem of detecting range-distributed targets in the presence of structured disturbance modelled as an autoregressive Gaussian process with unknown parameters. The focus is on two different scenarios. The first assumes that all the data vectors from the cells under test share the same covariance matrix (homogeneous environment). The second refers to the case of data vectors characterised by completely different covariances (heterogeneous environment). Four detectors exploiting the asymptotic generalised likelihood ratio criterion are devised and assessed. Remarkably, they ensure the constant false alarm rate (CFAR) property with respect to the disturbance power level, and two of them are asymptotically CFAR with respect to the disturbance covariance matrix. Finally the performance assessment, based also on real radar data, has shown that these detectors achieve, in general, satisfactory detection performances.
Article
Full-text available
A modified generalized likelihood ratio test (MGLRT) for the adaptive detection of a target or targets that are distributed in range is derived. The unknown parameters associated with the hypothesis test are the complex amplitudes in range of the desired target and the unknown covariance matrix of the additive interference, which is assumed to be characterized as complex zero-mean correlated Gaussian random variables. The target's or targets' complex amplitudes are assumed to be distributed across the entire input data block (sensor×range). Results on probabilities of false alarm and detection are derived, and a bounded constant false alarm rate (CFAR) detector is developed. Simulation results are presented. It is shown that the derived MGLRT of range distributed targets is much more effective in detecting targets distributed in range than an M out of K detector, which is cascaded with a single-point target Kelly (1986) detector
Article
Full-text available
A generalized likelihood ratio test (GLRT) is derived for adaptive detection of range and Doppler-distributed targets. The clutter is modeled as a spherically invariant random process (SIRP) and its texture component is range dependent (heterogeneous clutter). We suppose here that the speckle component covariance matrix is known or estimated thanks to a secondary data set. Thus, unknown parameters to be estimated are local texture values, the complex amplitudes and Doppler frequencies of all scattering centers. To do so, we use superresolution methods. The proposed detector assumes a priori knowledge on the spatial distribution of the target and has the precious property of having a constant false alarm rate (CFAR) with the assumption of a known speckle covariance matrix or by the use of frequency agility.
Article
Theoretical and applications aspects of high-spatial-resolution radars (HRRs) are examined in an introduction for upper-level engineering students. The advantages of increased radar bandwidths are discussed, and chapters are devoted to the application of the radar range equation to HRR, HRR design, waveforms and processing, synthetic high-range-resolution radar, SAR, inverse SAR, three-dimensional imaging with monopulse radar, target imaging with noncoherent radar systems, and applications to surveillance and countermeasures. Diagrams, graphs, a list of commonly used symbols, and sample problems with solutions are provided.
Article
The authors deal with the performance analysis of an adaptive version of the generalised matched subspace detector (GMSD) in compound-Gaussian clutter with unknown covariance matrix. The original GMSD was proposed to detect subspace signals in compound-Gaussian noise with known covariance matrix and ensures the constant false alarm rate (CFAR) property. In real situations, this assumption is unrealistic, which means that the covariance matrix must be estimated from training data. The authors use a robust estimate of the covariance matrix called the fixed-point estimate, recently proposed in the literature. The performance of the obtained adaptive detector, in terms of CFAR behaviour and probability of detection, is evaluated in the presence of real sea clutter data, collected by the McMaster IPIX radar.
Conference Paper
In this paper we present an analysis of sea clutter non-stationarity with respect to clutter covariance matrix estimation and its impact on the constant false alarm rate (CFAR) property of the normalized adaptive matched filter (NAMF). Three estimators have been considered in the analysis, the sample covariance matrix (SCM), the normalized sample covariance matrix (NSCM), and the fixed point (PF) estimators. The impact of non-stationarity, that emerges in the statistical analysis of the clutter data, is measured in terms of differences between NAMF nominal probability of false alarm (PFA) and probability of detection (PD) and quantified by processing measured clutter data recorded by the X-band IPIX radar.
Conference Paper
We use monostatic data obtained from the Multichannel Airborne Radar Measurements (MCARM) program to conduct a comparative analysis of various space-time adaptive processing techniques. Specifically, we consider reduced-dimension and reduced-rank methods. The measured data analysis shows the significance of training sample selection, effects of nonhomogeneity in the clutter data, sensitivity of certain classes of adaptive filters to the rank of the measured clutter signal and suitability of the various approaches.
Article
Adaptive detection of a sparsely distributed target is addressed without secondary data, in non-Gaussian clutter modelled as a spherically invariant random vector. By utilising different estimators of covariance matrix, an adaptive detection scheme is proposed for a sparsely distributed target, based on the generalised likelihood ratio test and the order statistics. Moreover, the adaptive detector with recursive estimator holds approximate constant false alarm rate property with respect to the clutter covariance matrix structure and the statistics of the texture. The performance assessment conducted by Monte-Carlo simulation confirms the effectiveness of the proposed detectors.
Article
This work addresses the problem of covariance matrix estimation for adaptive radar detection in correlated heavy tailed clutter. The clutter is modeled as a compound-Gaussian process with unknown statistics. An approximate maximum likelihood (AML) estimator is derived and compared to the maximum likelihood (ML) estimator; their calculation requires the iterative solution of a transcendental equation whose numerical convergence is obtained through the introduction of a constraint in the iteration. The performance of the two “constrained” algorithms is evaluated in terms of the Frobenius norm of the error matrix, of the computational complexity (i.e., the number of iterations), and of the constant false alarm rate (CFAR) property of the adaptive detector which makes use of them. Numerical results show that the AML estimator can be calculated with a very small number of iterations, it has a negligible performance loss with respect to the ML and less computational complexity. Finally, the AML guarantees the desired CFAR property to the detector.
Article
The problem of adaptive detection for spatially distributed targets in compound-Gaussian clutter is studied. We first derive the optimum NP detector and suboptimum two-step GLRT detector. For the two-step detection strategy, we also introduce three covariance matrix estimation strategies and evaluate their CFAR properties and complexity issues. Next, the numerical results are presented by means of Monte Carlo simulation strategy. In particular, the simulation results highlight that the performance loss due to adaptively estimating the texture is negligible, and that the loss due to adaptively estimating covariance matrix largely depends on the estimation algorithm, the number of the secondary data vectors and the number of the scatterers.
Article
The contribution of this paper is the derivation of the joint maximum likelihood (ML) estimator of complex amplitude and Doppler frequency of a radar target signal embedded in correlated non-Gaussian clutter modelled as a compound-Gaussian process. The estimation accuracy of the ML frequency estimator is investigated and compared with that of the well-known periodogram and ESPRIT estimators under various operational scenarios. The hybrid Cramer-Rao lower bound (HCRLB) and a large sample closed form expression for the mean square estimation error are also derived for Swerling I target signal. Finally, numerical results obtained by Monte Carlo simulation are checked by means of measured sea clutter data for the general case of fluctuating target amplitude.
Article
We address adaptive detection of a range-spread target or targets embedded in Gaussian noise with unknown covariance matrix. To this end, we assume that cells (referred to in the following as secondary data) that are free of signal components are available. Those secondary data are supposed to possess either the same covariance matrix or the same structure of the covariance matrix of the cells under test. In this context, we design detectors relying on the generalized likelihood ratio test (GLRT) and on a two-step GLRT-based design procedure. Remarkably, both criteria lead to receivers ensuring the constant false alarm rate (CFAR) property with respect to the unknown quantities. A thorough performance assessment of the proposed detection strategies, together with the evaluation of their processing cost, highlights that the two-step design procedure is to be preferred with respect to the plain GLRT. In fact, the former leads to detectors that achieve satisfactory performance under several situations of practical interest and are simpler to implement than those designed resorting to the latter
Article
In this paper, we consider the problem of adaptive detection for range-spread targets with known Doppler and unknown complex amplitude in compound Gaussian clutter. The speckle component of the clutter is modeled as an autoregressive (AR) process. By using the generalized likelihood ratio test (GLRT) approach, we will first estimate the AR parameters and the unknown complex amplitude, and then propose an adaptive AR-based GLR detector. The performance assessments are presented too. The computer simulations show that the proposed detector, without a priori information of the covariance matrix, has the same asymptotical performances as the two-step GLR-based detector with known covariance matrix.
Conference Paper
Discusses the use of high-resolution radar (HRR) to achieve an improvement in target detection. This is pursued by optimizing the transfer of energy from the transmitter via the target to the receiver according to the concept of matched transfer functions. In essence, the proposed technique makes use of a-priori information on the target EM scattering characteristics to condition the transmitted waveform and the receiver processing function
Article
This paper deals with the statistical modelling of radar backscattering from sea surface at low-grazing angles in high resolution radar systems. High-resolution polarimetric data at different range resolutions (60, 30, 15, 9 and 3 m) are analysed to highlight the differences in clutter statistical behaviour due to changes of resolution and/or polarisation. The clutter data were recorded by the IPIX radar of McMaster University in Grimsby, Ontario, Canada
Article
A detector for the case of a radar target with known Doppler and unknown complex amplitude in coloured noise of unknown covariance has been derived. The detector assumes that the noise is an autoregressive process and estimates the unknown parameters by maximum likelihood estimation for the use in the generalised likelihood ratio test. The asymptotic performance of this detector has been derived and it has been shown that for large data records, this detector is CFAR. By computer simulations it has been shown that for a moderate size of data record, the performance of this detector approaches the asymptotic results
Article
We design and assess adaptive schemes to detect extended and multiple point-like targets embedded in correlated Gaussian noise. Proposed algorithms rely on either the generalized likelihood ratio test (GLRT) or ad hoc procedures. Such detectors make it possible to get rid of distinct secondary data and guarantee the constant false alarm rate (CFAR) property with respect to the covariance matrix of the disturbance. A preliminary performance assessment, conducted by resorting to simulated data, also in comparison to the so-called modified GLRT (MGLRT) proposed in , has shown that newly introduced CFAR detectors may represent a viable means to deal with uncertain scenarios.
Article
A detector of a spatially distributed target in white Gaussian noise is developed. A reasonable distribution for the a priori target scatterer density is assumed, and a detector that incorporates this a priori knowledge is given. A simple detector form results, whose detection performance is robust over different scattering densities.
Article
A generalized likelihood ratio test (GLRT) for the adaptive detection of a target or targets that are Doppler-shifted and distributed in range is derived. The unknown parameters associated with the hypothesis test are the complex amplitudes in range of the desired target and the unknown covariance matrix of the additive interference, which is assumed to be characterized as complex zero-mean correlated Gaussian random variables. The target's or targets' complex amplitudes are assumed to be distributed across the entire input data block (sensor × range). The unknown covariance matrix is constrained to have the reasonable form of the identity matrix (the internal noise contribution) plus an unknown positive semidefinite (psdh) matrix (the external interference contribution). It is shown via simulation for a variety of interference scenarios that the new detector has the characteristic of having a bounded constant false alarm rate (CFAR), i.e., for our problem, the probability of false alarm PF for a given detection threshold is bounded by the PF that results when no external interference is present. It is also shown via simulation that the new detector converges relatively fast with respect to the number of sample vectors K necessary in order to achieve a given probability of detection PD
Article
The problem of detecting a known signal in colored Gaussian noise of unknown covariance is addressed. The noise is modeled as an autoregressive process of known order but unknown coefficients. By employing the theory of generalized likelihood ratio testing, a detector structure is derived and then analyzed for performance. It is proven that for large data records the detection performance is identical to that of an optimal prewhitener and matched filter, and therefore the detector itself is optimal. Simulation results indicate that the data record length necessary for the asymptotic results to apply can be quite small. Thus, the proposed detector is well suited for practical applications.
Article
The subject of detection of spatially distributed targets in non-Gaussian noise with unknown statistics is addressed. At the design stage, in order to cope with the a priori uncertainty, we model noise returns as Gaussian vectors with the same structure of the covariance matrix, but possibly different power levels (heterogeneous environment). We also assume that a set of secondary data, free of signal components, is available to estimate the correlation properties of the disturbance The proposed detector assumes no a priori knowledge about the spatial distribution of the target scatterers and ensures the constant false alarm rate (CFAR) property with respect to both the structure of the covariance matrix and the power levels. Finally, the performance assessment, conducted modeling the disturbance as a spherically invariant random process (SIRP), confirms its validity to operate in real radar scenarios
Article
A constant false alarm rate (CFAR) strategy for detecting a Gaussian distributed random signal against correlated non-Gaussian clutter is developed. The proposed algorithm is based on Scharf's matched subspace detector (MSD) and has the CFAR property with respect to the clutter amplitude probability density function (apdf), provided that the clutter distribution belongs to the compound-Gaussian family and the clutter covariance matrix is known to within a scale factor. Analytical expressions of false alarm and detection probabilities are derived. An application to the problem of detecting hovering helicopters against vegetated ground clutter is reported
Article
Two schemes for adaptive detection Kelly's generalized likelihood ratio test (GLRT) and the mean level adaptive detector (MLAD), are analyzed with respect to the deleterious effect of desired-signal contamination of the data used to compute the sampled covariance matrix for the two detectors. The detection probability P D and false alarm performance (ghosting probability P G) are predicted for the two schemes under the assumptions that the input noises are Gaussian random variables that are temporally independent but spatially correlated; and the amplitude of the desired signal is Rayleigh distributed. P D and P G are computed as a function of the false alarm probability P F with no contamination, the number of input channels, the number of independent samples-per-channel, the matched filtered output signal-to-noise (S/N) power ratio, and the S/N power ratio of the contaminating desired signal. It is shown that both P D and P G decrease with increasing levels of contamination. The P G performance is is almost identical for the GLRT and MLAD. The P D performance shows similar relative performance trends. Significantly, it is shown that the ghosting probability does not exceed P F in the presence of contamination.< >
  • S Haykin
  • A Steinhardt
S. Haykin, A. Steinhardt, Adaptive Radar Detection and Estimation, vol. 2, Wiley-Interscience, 1992.
  • S Kay
S. Kay, Fundamentals of Statistical Signal Processing, vol. 2, Prentice Hall, 1998.