The number of training data is usually limited for moving target detection in airborne radar, which can significantly degrade the performance of detectors. In this study, the authors propose a detector for detecting moving targets based on the random matrix theory (RMT). The clutter subspace is first estimated through the RMT. Then the data under test are projected onto the orthogonal complement space of the clutter subspace for whitening. Finally, the generalised energy accumulation detection of the whitened data is carried out. Simulation results show that the proposed detector can detect moving targets effectively even when the number of training data is extremely small and the detector has a fast rate of convergence and constant false alarm rate.

This study considers joint tracking and classification (JTC) of an extended object using measurements of down-range and cross-range extent. Using such measurements, existing approaches handle only tracking, that is estimating the kinematic state and the extension. In many practical applications, tracking and classification (e.g. classifying the object by its size and shape) are highly coupled (i.e. they affect each other) but are handled separately. For JTC of extended objects, this study deals with this problem jointly by integrating class-related extension information (i.e. the size and shape characteristics distinguishing objects of different classes) into a support function model. This facilitates the derivation of their JTC algorithm for jointly estimating the kinematic state and object extension and obtaining the probabilities of the object classes. In the proposed JTC algorithm, the useful information between the tracker and the classifier is sufficiently exchanged to improve overall performance. Furthermore, they also propose an effective method to fuse object extension estimates. The benefit of what they proposed is illustrated by simulation results.

Multi-baseline phase unwrapping operation is a very important step for multi-baseline Synthetic Aperture Radar Interferometry (InSAR). However, the conventional methods applied for single pixel suffer seriously from the phase noise that exists numerously in the interferometric phase images (i.e. interferograms). In order to improve the robustness to noise, this study proposes an innovative multi-baseline InSAR phase unwrapping method based on a mixed-integer optimisation model. The proposed method combines the central and its neighbouring pixels to jointly construct the mixed-integer optimisation model under the assumption that the pixels within a local window can be approximated by a small slant plane terrain. Furthermore, considering the practical case, the optimal window size is estimated according to the deviation from the interferometric wrapped phase to the assumed linear terrain, and the fast Fourier transform technique is adopted to reduce computational cost. The theoretical analysis and computer simulation demonstrate that the proposed method is able to improve the multi-baseline phase unwrapping performance and can be applied to reconstruct the digital elevation models for complicated topographies.

This study focuses on the problem of joint suppression of active jamming in target sector-of-interest (SOI) and out-of-sector interference for colocated compressive sensing multiple-input–multiple-output (CS-MIMO) radar. Three effective strategies for spatial filter measurement matrix (SFMM) design are outlined. Unlike the previous reported Capon beamformer and minimum variance distortionless response beamformer, the proposed design strategies only depend on the target spatial SOI rather than the accurate directions-of-arrival (DOAs) of targets, jammers and interfering sources to obtain deep nulls or notches for the SOI jamming and low-attenuation levels for the out-of-sector interference. The SFMM design criteria are derived using the second-order cone programming and solved as convex optimisation problems. The proposed approaches can simultaneously suppress the SOI jamming and attenuate the out-of-sector interference. Meanwhile, better DOA estimation accuracy can be achieved. Simulation results demonstrate the superiority of the proposed approaches over the other methods for colocated CS-MIMO radar.

To improve the sea-surface weak targets detection performance of the marine surface surveillance radar systems, the authors put forward a novel detection scheme based on a time-frequency distribution (TFD) fusion strategy assisted by population evolution algorithm. A Volterra-series-based weighted averaging model is utilised as the fusion rule to construct the fused TFD (FTFD), which aims to enhance the performance of time-frequency (TF) analysis and suppress signal-dependent cross-term artefacts. Herein, the optimal fusion coefficient is estimated by culture-based population evolutionary algorithm without any prior information. Unfortunately, this FTFD produces a great deal of redundant information. Hence, the normalised frequency marginal feature is extracted to reduce dimensions of the TF discriminant features, which is necessary to improve the efficiency of pattern classification. Finally, a multi-layered feed-forward neural network is utilised as a classifier in the pattern classification process. Experimental results demonstrate that the FTFD constructed by the proposed scheme achieves better performance in sharpness and strength than any subset of TFDs or their combinations and, furthermore, increases the detectability of sea-surface floating weak targets under any environment circumstances.

A joint processing framework is proposed for a distributed-phased multiple-input–multiple-output (phased-MIMO) sonar for detection of a small target. A transmit–receive-interactive strategy is combined with the array processing for a high signal-to-noise ratio (SNR) required by the distributed MIMO detection. The former provides the approximate location of a likely target by transmitting a probe signal to sense the environment. Then, the transmitting beams are steered to illuminate the target from different angles by which target echoes are enhanced and reverberation is suppressed. It is seen from the receiver operating characteristic curves that the distributed-phased-MIMO sonar is intermediate to the distributed MIMO sonar and the phased-array sonar in the low-SNR scenario. In the at-lake experiments of localisation of a small target, the distributed-phased-MIMO sonar system performs better than other two sonar systems due to exploiting the diversity gain and the transmitting array gain.

Human identification is crucial in various applications, including terrorist attack preventing, criminal seeking, defence and so on. Traditional human identification methods are usually based on vision, biological features, radio-frequency identification cards and so on. In this study, the authors propose an identification method based on radar micro-Doppler signatures using deep convolutional neural networks (DCNNs) for the first time, which can identify human in non-contact, remote and no lighting status. They employ a K-band Doppler radar to acquire the raw signals due to its stationary clutter rejection and movement detection ability as well as its short wavelength which can generate larger Doppler shift. Then short-time Fourier transform is applied to the raw signals to characterise micro-Doppler signatures. They adopt the DCNNs to deal with the spectrograms for human identification problem. The DCNNs can learn the necessary features and classification conditions from raw micro-Doppler spectrograms without employing any explicit features. While the traditional supervised learning techniques relying on the extracted features require domain knowledge of each problem. It is shown that this method can achieve average accuracy ∼97.1% for 4 people, 90.9% for 6 people, 89.1% for 8 people, 85.6% for 10 people, 77.4% for 12 people, 72.6% for 16 people and 68.9% for 20 people.

The latest works on the radar pulse de-interleaving in electronic support measures (ESM) systems have not completely solved the missing pulses problem yet. The authors present a pulse detection algorithm that aims at, if not eliminating it, diminishing the rate of pulses not detected because of the fact of being superimposed in other ones. Experiments set up on an ESM system based on software-defined radio (SDR) showed that the algorithm detection rate was near a hundred per cent and the false alarm rate was near zero when pulse amplitudes were >3.9 mV in the SDR input. This work is suggested for the areas of digital and radar signal processing and electronic warfare.

A novel tracking algorithm is proposed by the integration of the adaptive current statistical (CS) model and the modified strong tracking (ST) square-root cubature Kalman filter (SCKF) for the manoeuvring aircraft tracking problem. Firstly, the acceleration recursion equation and the acceleration mean input estimation are combined in order to realise the adaptive adjustment of the CS model. Then, the introduced position of the fading factor is relocated from the orthogonality principle and a new formula is put forward. Additionally, the strong manoeuver detection function is established to adjust the manoeuvring frequency of the CS model. The simulation results show that the proposed algorithm possesses better tracking accuracy than the multiple-fading-factor SCKF based on the CS model, the SCKF-ST filter based on the modified CS model and the interacting-multiple-model (IMM)-SCKF. Moreover, the proposed algorithm decreases the runtime by 40% compared with the IMM-SCKF.

A novel ultra wideband (UWB)-based scheme is proposed to provide robust and accurate robot localisation in indoor environments. An extended Kalman filter (EKF), which is suboptimal, is combined in the main estimator design with an extended unbiased finite impulse response (EFIR) filter, which has better robustness. In the integrated EKF/EFIR algorithm, the EFIR filter and the EKF operate in parallel and the final estimate is obtained by fusing the outputs of both filters using probabilistic weights. Accordingly, the EKF/EFIR filter output ranges close to the most accurate one of the EKF and EFIR filters. Experimental testing has shown that the EKF/EFIR-based UWB-range robot localisation is more robust than the EKF- and EFIR-based ones in uncertain noise environments.

This study examines space–time adaptive processing in the presence of non-independent and identically distributed (i.i.d.) clutter and array errors. The authors propose a clutter rank estimation method by exploring the spatial–temporal steering vectors of clutter. The proposed method is independent of clutter statistics and direction-independent array errors. They prove that when the proposed clutter rank estimation is used, the estimate of the clutter subspace is asymptotically independent of clutter statistics. This enables an eigensubspace method to acquire the asymptotic independence on clutter statistics. In addition, they prove that the eigensubspace method can suppress the clutter regardless of direction-independent array errors. They also suggest a geometrical non-homogeneity detector for the eigensubspace method. Simulation and experimental results with multi-channel airborne radar measurement (MCARM) data confirm that the eigensubspace method can suppress non-i.i.d. clutter such as discrete clutter as well as correlated clutter regardless of array gain-phase errors. The ability to suppress clutter regardless of clutter statistics and direction-independent array errors makes the eigensubspace method unique and feasible to the practical scenario when clutter is non-i.i.d. and the direction-independent array errors are present.

Synthetic aperture radar (SAR) imaging with curved trajectory is difficult due to its severe two-dimensional (2D) space variance. When processing the echo with azimuth-variance, the range cell migration and Doppler modulation of the targets with different azimuth locations are different, which brings a big challenge to image formation. Here, a 2D frequency decoupling method for this case is proposed. Firstly, the range history of the curved trajectory is accurately modelled based on fourth-order Taylor series expansion. Then, the 2D point target reference spectrum of the echo signal is obtained by the method of series reversion. The 2D space variance characteristics of two typical cases with curved trajectory, i.e. geosynchronous SAR and missile-borne SAR, are analysed. Based on the space variance characteristics, a 2D frequency decoupling step is implemented to handle the space-variant phase. Simulations are presented to demonstrate the effectiveness of the proposed algorithm.

An accurate fault detection method is critical in preventing the integrity of multi-source navigation system from the abnormal measurements which may occur any time. Here, a multi-channel single-dimensional fully convolutional neural network fault detection method is proposed, where the system measuring residuals sequence is used as the input, and the output is the system operating state, such as normal or fault types, in pointwise. The proposed technique extracts the features with various scales, which contain both the local and the general information of the signal sequence, for making a comprehensive and precise classification. To show the validity of the proposed method, computer simulations and trolley testing based on INS/GNSS/UWB integrated navigation system are carried out. The simulation and experimental results show that the proposed fault detection method is superior to the existing algorithms on the faults detection rate and false alarm rate, and thus, system reliability and navigation precision have been greatly improved.

Multipath (MP) is one of the main sources of errors in obtaining precise positioning using global navigation satellite systems (GNSSs) and continues to be extensively studied. In this study, a fast Fourier transform-based MP detection technique is proposed. The detector is formulated as a binary hypothesis test under the assumption that the MP exists for a sufficient time frame that allows its detection. The detection is based on the quadrature arm of the early-minus-late correlator output (Q _EmL) for a scalar tracking loop or on the quadrature (Q _EmL) and/or in-phase arm (I _EmL) for a vector tracking loop, using an observation window of N samples. Performance analysis of the proposed detector is done on multiple-ray (up to around 50) MP signals acquired from the MP environment simulator developed by the German Aerospace Centre (DLR in German). Both scalar and vector tracking schemes are used. The application of the detection test to exclusion of MP contaminated satellites from the navigation solution calculation significantly improves positioning accuracy as well as vector tracking performance. This detection technique can be extended to other GNSSs such as GLONASS, Galileo, and Compass with minor adjustments.