Igal Bilik

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.

Spatial Compressive Sensing for Direction-of-Arrival Estimation of Multiple Sources using Dynamic Sensor Arrays

This work addresses the problem of direction-of-arrival (DOA) estimation of multiple sources using short and dynamic sensor arrays. We propose to utilize compressive sensing (CS) theory to reconstruct the high-resolution spatial spectrum from a small number of spatial measurements. Motivated by the physical structure of the spatial spectrum, we model it as a sparse signal in the wavenumber-frequency domain, where the array manifold is proposed to serve as a deterministic sensing matrix. The proposed spatial CS (SCS) approach allows exploitation of the array orientation diversity (achievable via array dynamics) in the CS framework to address challenging array signal processing problems such as left-right ambiguity and poor estimation performance at endfire. The SCS is conceptually different from well-known classical and subspace-based methods because it provides high azimuth resolution using a short dynamic linear array without restricting requirements on the spatial and temporal stationarity and correlation properties of the sources and the noise. The SCS approach was shown to outperform current superresolution and orientation diversity based methods in single-snapshot simulations with multiple sources.

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.

MMSE-Based Filtering in Presence of Non-Gaussian System and Measurement Noise

The problem of sequential Bayesian estimation in linear non-Gaussian problems is addressed. In the Gaussian sum filter (GSF), the non-Gaussian system noise, the measurement noise, and the posterior state densities are modeled by the Gaussian mixture model (GMM). The GSF is optimal under the minimum-mean-square error (MMSE) criterion, however it is impractical due to the exponential model order growth of the system probability density function (pdf). The proposed recursive estimator, named the Gaussian mixture Kalman filter (GMKF), combines the GSF and the model order reduction procedure. The posterior state density at each iteration is approximated by a lower order density. This model order reduction procedure minimizes the estimated Kullback-Leibler divergence (KLD) of the reduced order density from the original density at each step. The estimation performance of the proposed GMKF is compared with the interactive multiple modeling (IMM), particle filter (PF), Gaussian sum PF (GSPF), and the GSF with mixture reduction (MR) method via simulations. It is shown in several examples that the proposed GMKF outperforms the other tested algorithms in terms of estimation accuracy. The superior estimation performance of the GMKF is obtained at the expense of its computational complexity, which is higher than the IMM and the MR algorithms.

A Single-Base-Station Localization Approach Using a Statistical Model of the NLOS Propagation Conditions in Urban Terrain

Implementation of the Federal Communications Commissions (FCC) E-911 Phase-2 requirements and commercial location-based services in urban environments requires mobile station (MS) localization capabilities in non-line-of-sight (NLOS) propagation conditions. This work addresses the problem of MS localization in urban environments characterized by the NLOS propagation conditions and limited Global Positioning System (GPS) services. We propose a single-base-station-based localization approach that uses a statistical model of urban propagation conditions in the framework of target classification. The proposed localization approach involves no data collection during the training process, requires no modifications of the MS hardware, is not location specific, and requires no identification and mitigation of the NLOS conditions. The performance of the proposed localization approach was evaluated using collected and ray-traced measurements.

Target tracking in glint noise environment using nonlinear non-Gaussian Kalman filter

The problem of nonlinear non-Gaussian target tracking with glint measurement noise is addressed in this work. The heavy-tailed glint noise distribution is modeled by mixture of two Gaussians. A new nonlinear Gaussian mixture Kalman filter (NL-GMKF), is applied to this problem. The tracking performance of the NL-GMKF is evaluated and compared to the particle filter (PF) and the extended Kalman filter (EKF) via simulations. It is shown that the NL-GMKF outperforms both the PF and the EKF.

Optimal recursive filtering using gaussian mixture model

Kalman filter is an optimal recursive estimator of the system state in terms of minimum-mean-square error (MMSE) under linear Gaussian assumptions. The Gaussianity assumption is not satisfied in many applications, such as dynamic channel estimation in mobile communications, maneuvering target tracking and speech enhancement. In this paper, the MMSE estimator for linear, non-Gaussian problems is presented, where the Gaussian mixture model is used for non-Gaussian distributions. The resulting recursive algorithm, named as non-Gaussian Kalman filter (NGKF), is composed of several conventional Kalman filters combined in an optimal manner. The performance of the proposed NGKF, is compared to the Kalman and particle filters via simulations. It is shown that the proposed NGKF outperforms both the Kalman and particle filters

Maneuvering Target Tracking Using the Nonlinear Non-Gaussian Kalman Filter

The problem of maneuvering target tracking is addressed in this paper. The main challenge in maneuvering target tracking stems from the nonlinearity and non-Gaussianity of the problem. The Singer model was used to model the maneuvering target dynamics and abrupt changes in the acceleration. According to this model, the heavy-tailed Cauchy distribution driving noise is used to model the abrupt changes in the target acceleration. The nonlinear, non-Gaussian Kalman filter was applied to this problem. The algorithm is based on the Gaussian mixture model for the posterior state vector. The nonlinear, non-Gaussian Kalman filter for this problem was tested using simulations, and it is shown that it outperforms both the particle filter and the extended Kalman filter

Requirements for next generation automotive radars

Active safety is one of the most dynamic emerging topics in the automotive industry. Radars currently play a major role in providing sensing capabilities required to meet active safety requirements. Requirements, parameters, and characteristics of automotive radars change as a function of the operation range. The main goal of this work is to summarize sensing requirements of both short- and long-range automotive radars and to highlight present challenges to both systems. This work discusses various aspects of automotive radars, states their main challenges, and suggests possible solutions.