Joseph Tabrikian

Target Detection and Localization Using MIMO Radars and Sonars

In this paper, we propose a new space-time coding configuration for target detection and localization by radar or sonar systems. In common active array systems, the transmitted signal is usually coherent between the different elements of the array. This configuration does not allow array processing in the transmit mode. However, space-time coding of the transmitted signals allows to digitally steer the beam pattern in the transmit in addition to the received signal. The ability to steer the transmitted beam pattern, helps to avoid beam shape loss. We show that the configuration with spatially orthogonal signal transmission is equivalent to additional virtual sensors which extend the array aperture with virtual spatial tapering. These virtual sensors can be used to form narrower beams with lower sidelobes and, therefore, provide higher performance in target detection, angular estimation accuracy, and angular resolution. The generalized likelihood ratio test for target detection and the maximum likelihood and Cramer-Rao bound for target direction estimation are derived for an arbitrary signal coherence matrix. It is shown that the optimal performance is achieved for orthogonal transmitted signals. Target detection and localization performances are evaluated and studied theoretically and via simulations

GMM-based target classification for ground surveillance Doppler radar

An automatic target recognition (ATR) algorithm, based on greedy learning of Gaussian mixture model (GMM) is developed. The GMMs were obtained for a wide range of ground surveillance radar targets such as walking person(s), tracked or wheeled vehicles, animals, and clutter. Maximum-likelihood (ML) and majority-voting decision schemes were applied to these models for target classification. The corresponding classifiers were trained and tested using distinct databases of target echoes, recorded by ground surveillance radar. ML and majority-voting classifiers obtained classification rates of 88% and 96%, correspondingly. Both classifiers outperform trained human operators.

Maximum a-posteriori probability pitch tracking in noisy environments using harmonic model

Modern speech processing applications require operation on signal of interest that is contaminated by high level of noise. This situation calls for a greater robustness in estimation of the speech parameters, a task which is hard to achieve using standard speech models. In this paper, we present an optimal estimation procedure for sound signals (such as speech) that are modeled by harmonic sources. The harmonic model achieves more robust and accurate estimation of voiced speech parameters. Using maximum a posteriori probability framework, successful tracking of pitch parameters is possible in ultra low signal to noise conditions (as low as -15 dB). The performance of the method is evaluated using the Keele pitch detection database with realistic background noise. The results show best performance in comparison to other state-of-the-art pitch detectors. Application of the proposed algorithm in a simple speaker identification system shows significant improvement in the performance.

Barankin bounds for source localization in an uncertain ocean environment

Ambiguity surfaces for underwater acoustic matched-field processing are prone to having high secondary peaks. This leads to anomalous source localization estimates below a threshold signal-to-noise ratio (SNR) at which the performance rapidly departs from that predicted by the Cramer-Rao lower bound (CRLB). In this paper, Barankin bounds are used to predict the threshold SNR under two different models including known or uncertain shallow-water environments and monochromatic or random narrowband sources. Evaluation of the Barankin bound suggests that although asymptotic localization performance degrades with increasing environmental uncertainty, the threshold SNR is relatively unaffected.

Relationships between adaptive minimum variance beamforming and optimal source localization

For many years, the popular minimum variance (MV) adaptive beamformer has been well known for not having been derived as a maximum likelihood (ML) estimator. This paper demonstrates that by use of a judicious decomposition of the signal and noise, the log-likelihood function of source location is, in fact, directly proportional to the adaptive MV beamformer output power. In the proposed model, the measurement consists of an unknown temporal signal whose spatial wavefront is known as a function of its unknown location, which is embedded in complex Gaussian noise with unknown but positive definite covariance. Further, in cases where the available observation time is insufficient, a constrained ML estimator is derived here that is closely related to MV beamforming with a diagonally loaded data covariance matrix estimate. The performance of the constrained ML estimator compares favorably with robust MV techniques, giving slightly better root-mean-square error (RMSE) angle-of-arrival estimation of a plane-wave signal in interference. More importantly, however, the fact that such optimal ML techniques are closely related to conventional robust MV methods, such as diagonal loading, lends theoretical justification to the use of these practical approaches.

Spatially coded signal model for active arrays

This paper addresses the problem of target detection and localization by radar or active sonar systems. A novel configuration, in which the transmitted signals are spatially coded, is proposed. The main advantages of this new configuration are: avoiding beam-shape loss, having a larger virtual array aperture and therefore narrower beams, increasing the angular resolution, and having the ability to detect and localize a greater number of targets. This configuration enables array processing in the transmit mode in addition to the receive mode. The generalized likelihood ratio test (GLRT) and the maximum-likelihood (ML) estimator are derived for target detection and localization according to the new model configuration. The performance of the array processing algorithms for this problem is studied theoretically and via simulations.

Optimal Adaptive Waveform Design for Cognitive MIMO Radar

This paper addresses the problem of adaptive waveform design for estimation of parameters of linear systems. This problem arises in several applications such as radar, sonar, or tomography. In the proposed technique, the transmit/input signal waveform is optimally determined at each step based on the observations in the previous steps. The waveform is determined to minimize the Bayesian Cramér-Rao bound (BCRB) or the Reuven-Messer bound (RMB) for estimation of the unknown system parameters at each step. The algorithms are tested for spatial transmit waveform design in multiple-input multiple-output radar target angle estimation at very low signal-to-noise ratio. The proposed techniques allow to automatically focusing the transmit beam toward the target direction. The simulations show that the proposed adaptive waveform design methods achieve significantly higher rate of performance improvement as a function of the pulse index, compared to other signal transmission methods, in terms of estimation accuracy.

Maneuvering Target Tracking in the Presence of Glint using the Nonlinear Gaussian Mixture Kalman Filter

The problem of maneuvering target tracking in the presence of glint noise is addressed in this work. The main challenge in this problem stems from its nonlinearity and non-Gaussianity. A new estimator, named as nonlinear Gaussian mixture Kalman filter (NL-GMKF) is derived based on the minimum-mean-square error (MMSE) criterion and applied to the problem of maneuvering target tracking in the presence of glint. The tracking performance of the NL-GMKF is evaluated and compared with the interacting multiple modeling (IMM) implemented with extended Kalman filter (EKF), unscented Kalman filter (UKF), particle filter (PF) and the Gaussian sum PF (GSPF). It is shown that the NL-GMKF outperforms these algorithms in several examples with maneuvering target and/or glint noise measurements.

Radar target classification using doppler signatures of human locomotion models

The problem of target classification for ground surveillance Doppler radars is addressed. Two sources of knowledge are presented and incorporated within the classification algorithms: 1) statistical knowledge on radar target echo features, and 2) physical knowledge, represented via the locomotion models for different targets. The statistical knowledge is represented by distribution models whose parameters are estimated using a collected database. The physical knowledge is represented by target locomotion and radar measurements models. Various concepts to incorporate these sources of knowledge are presented. These concepts are tested using real data of radar echo records, which include three target classes: one person, two persons and vehicle. A combined approach, which implements both statistical and physical prior knowledge provides the best classification performance, and it achieves a classification rate of 99% in the three-class problem in high signal-to-noise conditions.

An efficient vector sensor configuration for source localization

An electromagnetic vector-sensor enables estimation of the direction of arrival (DOA) and polarization of an incident electromagnetic wave with arbitrary polarization. In this letter, an efficient vector-sensor configuration is proposed. This configuration includes the minimal number of sensors, which enables DOA estimation of an arbitrary polarized signal from any direction except two opposite directions on the z-axis. The configuration is obtained by analyzing the Cramer-Rao lower bound (CRLB) for source localization using a single vector-sensor. The resulting vector-sensor configuration consists of two electric and two magnetic sensors. It is shown that this quadrature configuration satisfies the necessary and sufficient conditions for the DOA estimation problem. The CRLB for DOA estimation of signals in the azimuth plane is identical for the quadrature and the complete vector-sensor configurations.