Zheng Bao

Inverse synthetic aperture radar imaging of maneuvering targets

Range Doppler imaging, a basic method used in inverse syn- thetic aperture radar (ISAR), is based on the uniform rotation model of a target. The rotation velocity and axis of a maneuvering aircraft often vary with time. Traditional imaging methods may blur the target image and even render it beyond recognition. A concept is presented for range- instantaneous Doppler imaging and a corresponding imaging method that can clearly improve the image quality of most maneuvering aircraft. The resulting images of real data show that the method is effective.

Radar HRRP Statistical Recognition: Parametric Model and Model Selection

Statistical modeling for radar high-resolution range profile (HRRP) is a challenging task in radar HRRP statistical recognition. Theoretical analysis and experimental results show that elements in an HRRP sample are statistically correlated and non-Gaussian distributed. First, this paper introduces three joint-Gaussian models, i.e., subspace approximation model, probability principal components analysis (PPCA) model and factor analysis (FA) model, into radar HRRP statistical recognition. Due to the experimental results, we can have the conclusion that the jointly non-Gaussian distributed HRRP samples approximately follow the joint-Gaussian distribution described by FA model. Therefore, we can apply FA model to radar HRRP statistical recognition rather than a joint-Gaussian mixture model, e.g., PPCA mixture model or FA mixture model, which is a more accurate choice for modeling non-Gaussian distributed correlations in multidimensional data but with high learning complexity and large computation burden, and the difficulty in the statistical modeling for HRRP samples is largely reduced. Second, this paper concerns model selection of FA model in radar HRRP statistical recognition, in which there are two issues, i.e., the partition of target-aspect frames and the determination of the number of factors in each frame. Based on the Akaike information criterion (AIC) and the Bayes information criterion (BIC), an iterated algorithm for model selection is proposed in this paper, which can automatically give the optimal aspect-frame boundaries and determine the optimal number of factors in each aspect-frame. The recognition experiments based on measured data show that the proposed adaptive partition approach can further improve the recognition performance with higher recognition efficiency.

Properties of high-resolution range profiles

Wideband radar can obtain much target information due to its high-range resolution. Because its size is smaller than the resolution of conventional radar, for an airplane, it can be regarded as a point. However, for wideband radar with a bandwidth of several hundred megahertz, the range resolution is less than a meter, and airplane echoes of one impulse form a high-resolution range profile (HRRP), which includes the shape information of a target that can be used for automatic target recognition (ATR). A range profile can be divided into many range resolution cells, each of which still contains the echoes of many scatterers. The complex amplitude of one range cell echo can be regarded as the sum of these echoes. A small aspect of change will lead to a big change of the complex amplitude, since, although the variation of the range of each scatterer to radar is very small, it is still large enough compared to the small wavelength and may cause a big phase change. Hence, the range profiles are very sensitive to the aspect angle, which makes the ATR based on HRRPs a well-known challenging problem. The properties of range profiles are investigated and a preprocessing scheme is presented to obtain stable range profiles for ATR.

Bayesian Spatiotemporal Multitask Learning for Radar HRRP Target Recognition

A Bayesian dynamic model based on multitask learning (MTL) is developed for radar automatic target recognition (RATR) using high-resolution range profile (HRRP). The aspect-dependent HRRP sequence is modeled using a truncated stick-breaking hidden Markov model (TSB-HMM) with time-evolving transition probabilities, in which the spatial structure across range cells is described by the hidden Markov structure and the temporal dependence between HRRP samples is described by the time evolution of the transition probabilities. This framework imposes the belief that temporally proximate HRRPs are more likely to be drawn from similar HMMs, while also allowing for possible distant repetition or “innovation”. In addition, as formulated the stick-breaking prior and MTL mechanism are employed to infer the number of hidden states in an HMM and learn the target-dependent states collectively for all targets. The form of the proposed hierarchical model allows efficient variational Bayesian (VB) inference, of interest for large-scale problems. To validate the formulation, example results are presented for an illustrative synthesized dataset and our main application-RATR, for which we consider the measured HRRP data. For the latter, we also make comparisons to the model with the independent state-transition statistics and some other existing statistical models for radar HRRP data.

Radar HRRP Statistical Recognition With Local Factor Analysis by Automatic Bayesian Ying-Yang Harmony Learning

Radar high-resolution range profiles (HRRPs) are typical high-dimensional, non-Gaussian and interdimension dependently distributed data, the statistical modelling of which is a challenging task for HRRP based target recognition. Assuming the HRRP data follow interdimension dependent Gaussian distribution, factor analysis (FA) was recently applied to describe radar HRRPs and a two-phase procedure was used for model selection, showing promising recognition results. Besides the interdimensional dependence, this paper further models the non-Gaussianity of the radar HRRP data by local factor analysis (LFA). Moreover, since the two-phase procedure suffers from extensive computation and inaccurate evaluation on high-dimensional finite HRRPs, we adopt an automatic Bayesian Ying-Yang (BYY) harmony learning, which determines the component number and the hidden dimensionalities of LFA automatically during parameter learning. Experimental results show incremental improvements on recognition accuracy by three implementations, progressively from a two-phase FA, to a two-phase LFA, and then to an automatically learned LFA by BYY harmony learning.

Noise Robust Radar HRRP Target Recognition Based on Multitask Factor Analysis With Small Training Data Size

A factor analysis model based on multitask learning (MTL) is developed to characterize the FFT-magnitude feature of complex high-resolution range profile (HRRP), motivated by the problem of radar automatic target recognition (RATR). The MTL mechanism makes it possible to appropriately share the information among samples from different target-aspects and learn the aspect-dependent parameters collectively, thus offering the potential to improve the overall recognition performance with small training data size. In addition, since the noise level of a test sample is usually different from those of the training samples in the real application, another contribution is that the proposed framework can update the noise level parameter in the FA model to adaptively match that of the received test sample. Efficient inference is performed via variational Bayesian (VB) for the proposed hierarchical Bayesian model, and encouraging results are reported on the measured HRRP dataset with small training data size and under the test condition of low signal-to-noise ratio (SNR).

On the aspect sensitivity of high resolution range profiles and its reduction methods

It has been shown that high resolution radar range profiles (HRRP) can serve as discriminative features for automatic target recognition (ATR) purpose. A HRRP at a single aspect, however, is very sensitive to the aspect change. For microwave radars, aspect change of tenths of 1/spl deg/ can cause drastic changes in HRRP of aircraft-like targets. The aspect sensitivity of HRRP greatly hinders their potential use in automatic target recognition. This paper investigates the phenomenon of aspect sensitivity of HRRP and methods for its reduction.

Nonparametric Detection of FM Signals Using Time-Frequency Ridge Energy

In many practical applications, signals to be detected are unknown nonlinear frequency modulated (FM) and are corrupted by strong noise. The phase histories of signals are assumed to be unknown smooth functions of time and these functions are poorly modeled or unmodeled by a small number of parameters. Thus, the conventional parametric-based detection methods are invalid in these cases. This paper proposes a nonparametric detection method using the ridge energy of observations. The detection process consists of three steps, TF ridge detection, ridge energy extraction, and decision. First, the directionally smoothed-pseudo-Wigner-Ville distribution (DSPWVD) is introduced to highlight the instantaneous frequency (IF) points along a special direction on the IF curve of a signal from noise. Further, an angular maximal distribution (AMAD) is constructed from a set of DSPWVDs to highlight the entire IF curve. As a result, the TF ridge of an observation can be estimated well from its AMAD by the maxima position detector. Second, the ridge energy, the total energy along the TF ridge on the pseudo-Wigner-Ville distribution (PWVD), is extracted. A noisy signal has larger ridge energy than a pure noise does, with a large probability, because pure noise energy is randomly distributed throughout the TF plane while the signal energy in a noisy signal is concentrated along the estimated TF ridge. Third, the ridge energy of an observation is used as the test statistic to decide whether or not a signal of interest is present in the observation, where the decision threshold is determined by a large number of Monte Carlo simulations using pure noise. Finally, the simulation experiments to two test signals are made to verify the effectiveness of the proposed method.

Range-Spread Target Detection Based on Cross Time-Frequency Distribution Features of Two Adjacent Received Signals

High resolution radars (HRRs) transmit a wideband signal to achieve a high range resolution. A target is considered as composed of multiple scatterers, which occupy or spread in multiple radar range cells with several scatterers in each cell. Therefore, the reflection of a target spreads in multiple range cells in the received signal, which contains more information of target than that obtained from low resolution radars. The target in high resolution radar systems is a range-spread target. The range-spreading or echo features of target are utilized for target detection and identification. The echoes of target are convolutions of transmitted signals with target range-scattering functions dependent on the gesture of target to the line of radar sight. A single echo is used in the conventional detection. It is difficult for target detection and identification in low signal-to-noise ratio (SNR) condition. In this paper, we propose a new range-spread target detection scheme exploiting the image features of cross time-frequency distribution (TFD) of a pair of adjacent received signals. After dechirping, the received signal reflected from target consists of multiple sinusoidal components due to its multiple scatterers when a linear frequency modulated (LFM) signal is transmitted from radar. Some regular image patterns or features of target appear in the cross TFD of two adjacent received signals, while the cross TFD of two independent Gaussian noises does not show such patterns. The cross TFD features are exploited in the proposed scheme. Three steps are composed in the proposed scheme. Firstly, a cross smoothed-pseudo Wigner-Ville distribution (CSPWVD) is made for two adjacent received signals to generate a two-dimensional (2-D) TF image. Then, some regular geometric patterns are detected and extracted from the image. At last, two features of the extracted geometric patterns are jointly utilized to detect target. The proposed algorithm is verified by using raw radar data. It outperforms the conventional detection methods.

A neural network learning for adaptively extracting cross-correlation features between two high-dimensional data streams

This paper proposes a novel cross-correlation neural network (CNN) model for finding the principal singular subspace of a cross-correlation matrix between two high-dimensional data streams. We introduce a novel nonquadratic criterion (NQC) for searching the optimum weights of two linear neural networks (LNN). The NQC exhibits a single global minimum attained if and only if the weight matrices of the left and right neural networks span the left and right principal singular subspace of a cross-correlation matrix, respectively. The other stationary points of the NQC are (unstable) saddle points. We develop an adaptive algorithm based on the NQC for tracking the principal singular subspace of a cross-correlation matrix between two high-dimensional vector sequences. The NQC algorithm provides a fast online learning of the optimum weights for two LNN. The global asymptotic stability of the NQC algorithm is analyzed. The NQC algorithm has several key advantages such as faster convergence, which is illustrated through simulations.