Micro-Doppler signals refer to Doppler scattering returns produced by the motions of the target other than gross translation. The small micro-motions of a subject, and even just parts of a subject, can be observed through the micro-Doppler signature it creates in response to an active emitter such as a radar, laser, and even acoustic emitters. These micro-Doppler signatures are produced by the kinematic properties of the subject's motion and can be used to extract the salient features of the subject's motion, and often, identify the subject. The rapidly declining cost of micro-Doppler-capable active sensors like radar with their dramatically improving capabilities, provide significant motivation in developing micro-Doppler techniques that can improve the exploitation of these sensors. Micro-Doppler techniques aim at extracting the micro-motion of the subject that may be unique to a particular subject class or activity in order to distinguish probable false alarms from real detections, as well as to increase the value of the information extracted from the sensor. The source of micro-motion depends on the subject and can be a rotating propeller on a fixed-wing aircraft, the multiple spinning rotor blades of a helicopter, or an unmanned aerial vehicle (UAV); the vibrations of an engine shaking a vehicle; an antenna rotating on a ship; the flapping wings of birds; the swinging arms and legs of a walking person; and many other sources. Confuser detections, such as birds for UAVs or animals for humans, can be interpreted as false alarms for a sensor system, so using the available micro-Doppler returns for classification can significantly reduce the sensor false alarm rate, thereby improving the utility of the sensor system. This study reviews the current research in micro-Doppler based on subject type, sensor capabilities, as well as environmental effects, and then proposes future research areas for micro-Doppler.

In radar imaging, the micro-Doppler effect is caused by fast movements of some scattering points on the target. These movements correspond to highly non-stationary components in the time–frequency domain of the signal. The rigid body can be considered as stationary at one range location during the processing time. This property is used to separate the micro-Doppler signal from the rigid body using the L-statistics. Since the rigid body can be considered as a sparse signal, its values can be fully recovered at the positions where the micro-Doppler and rigid body components overlap. The recovery is based on the compressive sensing theory and methods. After an overview of the methods, a quantitative analysis of the improvements achieved in the time–frequency-based separation is done. Moreover, a comparison with both the time and the frequency domain analysis is provided. Analysis of small additive noise influence to the reconstruction accuracy is done.

Micro-Doppler effect corresponds to non-stationary components in the time–frequency domain, while the rigid body can be considered as a stationary signal during the processing time. This property is used in the first part of the study to present the method for separation recovery based on the compressive sensing theory. Recovery bounds are based on the restricted isometry property or the coherence and spark analysis. Their calculation is a computationally difficult (Non-deterministic Polynomial-time (NP) hard) problem. These bounds also produce very pessimistic results. It is the reason to consider the statistical analysis of the reconstruction results. The statistical approach is presented in this part of the study and illustrated by examples, including a detailed case study example.

Inverse synthetic aperture radar (ISAR) imaging plays an important role in aerial target identification. However, traditional ISAR methods are based on rigid body dynamics, which cannot depict micro-moving parts, such as the rotating propellers of an aircraft. In this study, a novel wideband radar imaging algorithm is proposed to identify rapidly rotating objects. The proposed algorithm contains two key steps: the first involves the establishment of the algebraic equations based on the high-resolution range profile series. The second step consists of a solution for these equations, for which the authors provide the steepest descent iterative method as well as several improvements to enhance imaging quality. A few necessary preprocesses and potential challenging scenarios are also considered. The results are used from simulations and field data to exhibit the effectiveness and advantages of the proposed algorithm.

An approach is examined for reducing false alarms during detection of human targets via radar surveillance from moving platforms. The ability to distinguish between human target and clutter discrete micro-Doppler signatures is quantified using simulated and real data collected using the DORC X-band wideband experimental airborne radar. Signal processing methodologies to extract key target signature characteristics are discussed along with separability of target classes and classifier performance.

Safety and security applications benefit from better situational awareness. Radar micro-Doppler signatures from an observed target carry information about the target's activity, and have potential to improve situational awareness. This article describes, compares, and discusses two methods to classify human activity based on radar micro-Doppler data. The first method extracts physically interpretable features from the time-velocity domain such as the main cycle time and properties of the envelope of the micro-Doppler spectra and use these in the classification. The second method derives its features based on the components with the most energy in the cadence-velocity domain (obtained as the Fourier transform of the time-velocity domain). Measurements from a field trial show that the two methods have similar activity classification performance. It is suggested that target base velocity and main limb cadence frequency are indirect features of both methods, and that they do often alone suffice to discriminate between the studied activities. This is corroborated by experiments with a reduced feature set. This opens up for designing new more compact feature sets. Moreover, weaknesses of the methods and the impact of non-radial motion are discussed.

In recent years, Doppler radar has been used as a sensing modality for human gait recognition, due to its ability to operate in adverse weather and penetrate opaque obstacles. Doppler radar captures not only the speed of the target, but also the micro-motions of its moving parts. These micro-motions induce frequency modulations that can be used to characterise the target movements. However, a major challenge in Doppler signal processing is to extract discriminative features from the radar returns for target classification. This study presents a feature extraction method for classification of human motions from the micro-Doppler radar signal. The proposed method applies the log-Gabor filters at multiple spatial frequencies and orientations on a joint time–frequency representation. To achieve invariance to the target speed, features are extracted from local patches along the torso Doppler shift. Then, the (2D)²PCA (two-directional two-dimensional principal component analysis) method is applied to create a compact feature vector. Experimental results based on real radar data obtained from multiple human subjects demonstrate the effectiveness of the proposed approach in classifying arm motions.

A key challenge for radar surveillance systems is the discrimination of ground-based targets, especially humans from animals, as well as different types of human activities. For this purpose, target micro-Doppler signatures have been shown to yield high automatic target classification rates; however, performance is typically only given for near-optimal operating conditions using a fixed set of features. Over the past few decades dozens of micro-Doppler features have been proposed, when in fact utilisation of all possible features does not guarantee the maximum classification performance and the selection of an optimal subset of features is scenario dependent. In this work, a comprehensive survey of micro-Doppler features and their dependence upon system parameters and operational conditions – such as transmit frequency, range and Doppler resolution, antenna–target geometry, signal-to-noise ratio, and dwell time – is given. Algorithms for optimising classification performance for a reduced number of features are presented. Performance gains achievable using adaptive feature selection are assessed for a case study of interest.

This study presents the results of the authors’ experimental investigation into the radar micro-Doppler signatures (MDS) of various human activities both in free-space and through-wall environments. The collection of MDS signatures was divided into three categories: stationary, forward-moving, and multi-target. Each category of MDS signatures encompassed a variety of movements associated with it, adding up to a total of 18 human movements. Using a 6.5-GHz C-band coherent radar, the MDS of six human subjects were gathered in free-space and through-wall environments. The MDS for these cases were analysed in detail and the general properties of the signatures were related to their associated phenomenological characteristics. Based upon the MDS, features for designing detectors and classifiers of human targets performing such movements are recommended. In the case of multiple human targets in the radar field of view, it was found that it is possible to distinguish these targets from the MDS under certain circumstances, but not under others.

In this study, a feature extraction algorithm is presented which automatically generates a set of shape spectrum features based on the cadence velocity diagram of the human micro-Doppler signature. Recognition performance between humans undertaking the same activity is assessed on a set of experimental data collected with a continuous wave radar operating at X-band using a Naïve Bayesian classifier and a shape-similarity-spectrum classifier. Recognition performance is analysed as a function of key parameters, such as the dwell time on the target and the size of the training set, to investigate the level of robustness of the proposed features. Results show that high level recognition performance can be achieved for both the walking and running activities.

In this study, the authors consider the problem of human gait recognition in the presence of a walking cane using radars. Quadratic time–frequency distributions are used to provide the local signal behaviour over frequency and to detail the changes in the Doppler and micro-Doppler signatures over time. New features that capture the intrinsic differences in the time–frequency signatures of the gait observed with and without the use of a cane are proposed. The results based on real data experiments conducted in a laboratory environment are provided that validate the effectiveness of the proposed features in discriminating gait with cane from normal human gait.

This study discusses the analysis of multistatic micro-Doppler signatures and related features to distinguish and classify unarmed and potentially armed personnel. The application of radar systems to distinguish different motion types has been previously proposed and this work aims to further investigate the applicability of this in more scenarios. Real data have been collected using a multistatic radar system in a series of experiments involving several individuals performing different movements. Changes in classification accuracy as a function of different aspect angle between the direction in which the target faces and the line-of-sight of the radar nodes are analysed. Multiple data fusion methodologies are proposed, showing that significant improvement of the classification accuracy can be achieved when using separate classification at each node followed by a voting procedure to reach the final decision. This is beneficial especially at those aspect angles for which micro-Doppler detection is less favourable.

A novel concept, called range-Doppler surface (RDS), for human target analysis using ultra-wideband radar is proposed. The construction of RDS involves range-Doppler (RD) imaging, adaptive threshold detection and isosurface extraction. A Keystone-transform-based range migration compensation approach is proposed to allow high-quality RD imaging using ultra-wideband radar. Adaptive threshold detection is applied to detect the extended target in the RD image, and RDS is constructed by extracting an isosurface from a RD video sequence, which is defined as a sequence of RD images. In comparison with micro-Doppler profiles and high-resolution range profiles, RDS contains range, Doppler and time information simultaneously. An ellipsoid-based human motion model is designed for validation. RDSs simulated for different human activities are demonstrated and discussed. Finally, experimental results for single/two-people walking scenarios are presented to verify the simulation results. The use of the RDS opens a new area of human target analysis.

Extraction of features and subsequent classification of ground moving targets, especially humans and vehicles, are topics of great relevance for the theoretical research and practical application in the signal processing of the ground surveillance radar. Through the time–frequency analysis of ground moving target, the energy distribution of spectrogram can be regarded as a kind of image texture. On the basis of this, a novel method for the feature extraction of micro-motion targets is proposed in this study. In the proposed method, the spectrograms of targets’ echoes are first obtained through the short-time Fourier transform. To improve the precision of features extracted, a pre-processing based on the spectrogram is followed to enhance image features. On the basis of targets’ spectrograms, the entropy, third-order moment of statistical histogram and directionality features are extracted finally as jointly effective features of micro-motion targets. The support vector machine (SVM) is utilised to classify the ground moving targets and a high probability of correct classification is obtained. The experimental results under different signal-to-noise ratio and training sample number conditions verify the validity and robustness of the proposed method.

Time–frequency (TF) distributions have been used for providing high-resolution representation in a large number of signal processing applications. However, high resolution and accurate instantaneous frequency (IF) estimation usually depends on the employed distribution and complexity of signal phase function. To ensure an efficient IF tracking for various types of signals, a class of complex-time distributions (CTD) has been developed. These distributions facilitate analysis in cases when standard distributions cannot provide satisfactory results (e.g. for highly non-stationary signal phase). In that sense, an ambiguity-based form of the fourth-order CTD is considered, in a new compressive sensing (CS) context. CS is an intensively growing approach in signal processing that allows efficient analysis and reconstruction of randomly under-sampled signals. In this study, randomly chosen ambiguity domain coefficients serve as CS measurements. By exploiting sparsity in the TF plane, it is possible to obtain highly concentrated IF using just small number of randomly chosen coefficients from the ambiguity domain. Moreover, in noisy signal case, this CS approach can be efficiently combined with the L-statistics producing robust TF representations. Noisy coefficients are first removed using the L-statistics and then reconstructed by using the CS algorithms. The theoretical considerations are illustrated using experimental results.

An application of machine intelligence technique for the identification of micro-Doppler features from an airborne pulsed-Doppler radar sensor is developed. The key challenges for surveillance mode are the dynamic nature of the wind farm clutters, short-CPI length, and lack of prior information on the specific wind turbine (WT) in the site. The micro-Doppler spectrum segments based on short CPIs are used as the fundamental feature vectors for detection and classification. Both supervised and unsupervised approaches, including artificial neural network and random forest, are applied to airborne plan position indicator scan outputs. A simulator for airborne pulsed-Doppler radar operation over wind farm is used with realistic WT scattering signatures, platform motion impacts as well as the terrain clutter impacts. Based on the clutter identification result, the feasibility of detecting small moving targets in the presence of WT clutter is discussed.

This study presents an analysis of micro-Doppler signatures of helicopters obtained using a network of passive radar receivers utilising multistatic geometry. In the multistatic scenario, for each passive radar receiver the target is visible from different bistatic angles, which results in different micro-Doppler signatures. This feature allows additional information on the observed helicopter target to be obtained. The presented concept has been successfully verified by the authors using the real micro-Doppler signatures of a helicopter collected during the measurement campaign. As an illuminator of opportunity for the passive radar network a commercial digital video broadcasting-terrestrial (DVB-T) transmitter was used. The results of the applied micro-Doppler analysis for the real multistatic passive radar measurement carried out are presented in this study. In addition, some examples of inverse synthetic aperture radar (ISAR) images of helicopter rotor blades based on micro-Doppler analysis in the multistatic configuration are shown in this study.

Effective detection of marine targets with low observability gives a severe challenge to radar signal processing. The micro-Doppler (m-D) signatures of marine target can provide extra information for non-stationary and time-varying signal analysis. Long-time integration is an effective way to strengthen the m-D signal and improve signal-to-clutter ratio. However, the performances are affected by the range across unit and Doppler frequency migration effects. In this study, m-D characteristics of marine target are studied and a novel representation, i.e. phase differentiation and Radon-Lv's distribution (PD-RLVD), is proposed to detect the m-D signal to realise the long-time coherent integration. The PD-RLVD can accurately and directly represent the m-D signal in the chirp rate and chirp change domain appearing as obvious peaks. The proposed method is simple not only because it only requires a 2D Fourier transform of the scaled Radon instantaneous auto-correlation function after PD, but also for not introducing any non-physical attributes. Relations between PD-RLVD and other integration methods are introduced as well. Experiments with real data show that the proposed method can achieve higher integration gain, detection probability, and accuracies of motion parameter estimation.

The estimation of breathing activity of laboratory animals after drug administration attracts considerable interest in the context of the pharmacological experimentation. So far, this task has been mostly accomplished by means of expensive and cumbersome procedures requiring the application of sensors on the animal body. In this study, the authors present a feasibility study on the possible usage of bioradar devices for contactless monitoring the respiratory rhythm of living rodents. Experiments are performed in laboratory conditions on sleeping rats by using a continuous wave Doppler radar operating at 13.8 GHz. The recorded signals are processed by means of a data processing strategy, based on a novel motion artefacts filtering procedure, with the final aim to characterise the breathing pattern variability during the sleep phases. The achieved results are consistent with biological information available from the literature confirming the potential application of bioradiolocation instead of standard on-body monitoring methods.

Multiple-input–multiple-output (MIMO) technology has been suggested as an effective tool in overcoming some of the issues that are typically relate to conventional over the horizon radars. Notwithstanding, effects, such as fading and multipath propagations are ever present and cannot be avoided even when using a MIMO configuration. For this reason, a study on the impact of such effects on high frequency (HF) skywave MIMO radars is fundamental for an effective design of such systems. This study aims to study the effects of ionospheric propagation on the performance of the HF MIMO skywave radar. In particular, the relationship between the transmitted signal parameters and the ionospheric variations because of perturbations is highlighted in a suitable signal model. The performance analysis is performed in terms of estimated direction of arrival and loss of virtual array elements.

By exploiting the intrinsic sparsity of the spatial spectrum for a given resolution cell of range and Doppler, this study introduces the recently developed sparse representation approaches for direction-of-arrival estimation in shipborne high-frequency surface wave radar (HFSWR). These approaches can reconstruct the sparse signal and obtain high-resolution spatial spectrum with a small number of snapshots or even single snapshot. A generalised real-valued sparse representation, which seeks to replace the complex-valued problem with a real one, is proposed to reduce the computational complexity and improve the resolution probability and the estimation accuracy. The advantage of the proposed method is demonstrated by theoretical analysis and numeral simulations. Furthermore, validation of the proposed method is verified by the experiment of shipborne HFSWR on the Yellow Sea of China.

Range migration is a well-known, and readily compensable, phenomenon in synthetic-aperture radar. Under certain circumstances, it may also afflict phased array radar systems based on fixed antenna elements, but there is little or no literature on its occurrence in this type system, nor how it may be addressed. This study sets out to define the conditions in which range migration is a significant issue in phased array radar systems and how it may be mitigated, in particular concentrating on step frequency and frequency-modulated continuous-wave radar for which range migration is not so easily addressed. Results are presented to validate the approach based both on simulation and experimental data from ground-penetrating radar systems performing two-dimensional through-ice imaging in Antarctica.

This study develops a two-scan data association algorithm for the hybrid target state, which consists of the target trajectory state and the target existence state. The proposed method uses the sliding window-based approach for the multi-scan data association. It updates the target existence state by considering all the feasible multi-scan target existence events in the sliding window. The proposed method also updates the target trajectory state by calculating the joint association events in the sliding window. Performance of the proposed algorithm is compared with that of the integrated track splitting algorithm and the integrated probabilistic data association algorithm for various conditions.

This study investigates the problem of rapid detection of linear frequency modulated signals. An accelerating search algorithm is proposed based on the classical dechirp method. The fast Fourier transform operation is substituted with the truncated convolution in the dechirp method, and then a compatible search strategy is presented. The computational load of the proposed method is reduced. With the increasing length of the input signal, the presented method enhances the computational efficiency more obviously. Both the theoretical analysis and simulation results show that the detection performance of this method is desirable in the case of low signal-to-noise ratio; thus, it could be a competitive candidate for applications in engineering practices.

It is presumed that using multiple antennas at the transmit and receive sides can improve the radar's performance. One such improvement can be achieved in the detection of extended targets. However, in such multiple-input multiple-output systems, proper design of the transmit signals is an obstacle that can have significant influence on the performance. In this study, enforcing the peak-to-average-ratio (PAR) constraint, the authors consider the problem of transmit space-time code and receive space-time filter design of such a system detecting extended targets. In addition, the power allocation among the different antennas is another degree of freedom that should be optimised under a total power constraint. Here, the authors derive a constrained optimisation procedure for transmit code, receive filter and transmit powers, which sequentially improves the signal-to-interference plus noise ratio of the system.

The use of spaceborne synthetic aperture radar (SAR) for ground moving target indication (GMTI) has recently attracted a great deal of interest. In this study, an efficient imaging algorithm for spaceborne SAR/GMTI systems is proposed. The proposed imaging algorithm first implements the secondary range compression and the range cell migration correction in the two-dimensional frequency domain and then the azimuth compression in the range Doppler domain. The proposed algorithm can finely focus the stationary scene and coarsely focus multiple slow-moving targets simultaneously without a priori knowledge of targets’ motion parameters and position parameters. Moreover, the proposed algorithm is interpolation free and can be efficiently implemented by using only complex multiplications and fast Fourier transforms. These advantages make the proposed algorithm useful in practice. Experimental results from both simulated data and real data validate the proposed algorithm.

The micro-Doppler (m-D) signature induced by the micro-motion dynamics of a component or structure on a radar target can provide additional information for classification, recognition, and identification of the object. Since the m-D signature from conventional synthetic aperture radar (SAR) is insufficient for further application, attention centred on multi-input–multi-output (MIMO) SAR. MIMO SAR employs diversity to improve SAR performance and shows great potential in applications including high-resolution wide-swath remote sensing, slow moving target detection, and three-dimensional (3D) imaging. Take vibration for research, because vibration is a basic kind of micro-motion and it occurs commonly such as vehicle engine. In this study, the vibration signal modelling is deduced and the m-D signature of vibrating target is analysed, based on short-term shift-orthogonal waveform in MIMO SAR. Simulations verify the influence of vibration parameters on imaging and time–frequency profile and show the m-D signature difference through different MIMO channels. The results demonstrate the potential of MIMO SAR for 3D micro-motion estimation.

Aiming at interferometric inverse synthetic aperture radar (InISAR) imaging on squint model, a new method based on respective reference range selection (RRRS) and squint iteration improvement (SII) is proposed in this study. The influence of squint model mainly includes two aspects, namely accessional phase and image distortion. The three-dimensional (3D) imaging results are degraded greatly with the increment of squint angle by using traditional InISAR imaging algorithm. On the basis of the RRRS, the image registration is well achieved. Then, after a few times of SII and distortion correction, the performance of 3D reconstruction was improved. Simulation results confirmed the effectiveness of the proposed method.

This study proposes a novel non-negative matrix factorisation (NMF) variant L _1/2-NMF after visualisation and analysis of the process of target recognition via NMF for synthetic aperture radar (SAR) images. NMF has been applied to obtain pattern feature in SAR images. This study considers the intrinsic character and the physical meaning of NMF feature when applied for SAR automatic target recognition. At the base of obtaining the linear relationship between the sample to be recognised and the train samples, the whole recognition process is detailed and vividly visualised. Meanwhile, lots of researches have been done to improve NMF methods by enforcing sparse constraint with L ₁-norm, such as non-negative sparse coding (NNSC), local NMF and sparse NMF. Compared with L ₁-norm, L _1/2-norm has been shown to have a more natural sparseness. In this study, a novel variant of NMF with L _1/2 constraint, called L _1/2-NMF is proposed, and is carried out a thorough study by applying it in SAR target recognition. Experimental results on MSTAR public database show that both the basis and coding matrices obtained by L _1/2-NMF have higher sparseness than those obtained by NMF, NNSC and NMF with sparseness constraints (NMFsc). The recognition results demonstrate that the L _1/2-NMF outperforms NNSC, NMFsc and non-smooth NMF.