The linear frequency modulated (LFM) frequency modulated continuous wave (FMCW)-based frequency diverse array (FDA) radar concept is investigated in detail. The radar operates as a linear pulsed FMCW/FDA in the transmission (TX) mode while it operates as a pulsed FMCW/phased array (PA) in the receiving mode. The issues such as low signal-to-noise ratio (SNR) of FDA, the time-angle scanning and time-range ambiguities are studied. It is shown that the local instantaneous frequency bandwidth is much smaller than the radio frequency (RF) deviation of LFM. Positive and negative slope TX/RF locations offer frequency diversity. Time domain and frequency domain signal processings are described. A Ku band direct digital synthesis-based FMCW/FDA radar example based on the cumulative detection scheme is given and compared with an equivalent FMCW/PA radar.

Inter-radar interference will be a crucial problem when a number of frequency modulated continuous wave (FMCW) radars are densely deployed in future, since it can cause miss detection and false detection of the target due to wideband interference and narrowband interference. One of the approaches to solve this problem is to introduce a multiple access technique for automotive radars, e.g. carrier sense multiple access used in wireless LAN to avoid packet collision. This study proposes multiple access FMCW radar which can avoid both narrow-band and wideband interference assuming all of the FMCW radars have the same design parameters. It also, proposes a carrier sense technique as a key technique for multiple access FMCW radar and evaluates the carrier sense performance by computer simulations.

Automatic targets recognition (ATR) for synthetic aperture radar (SAR) image is very important. ATR can be used in traffic management, national frontier safety, and so on. Traditional algorithms for SAR ATR is composed of extraction features and training classifier. The features are essential for the classification accuracy. However, choosing good features by hand is a hard task. The deep convolutional neural networks (CNNs) which can learn features automatically have got a great performance in natural images. However, the CNNs have many parameters and need a lot of data to train such networks. The remote-sensing data of SAR is limited. Then, the authors need a simple network which needs not much data and easy to train. The principal component analysis network (PCANet) is a shallow network that performs well in the recognition task and needs no hand features choosing. Though this network has produced a wide application in the natural images, it is rarely used in the SAR images. The experimental result of the moving and stationary target acquisition and recognition (MSTAR) dataset shows that the PCANet can achieve over 99% accuracy on ten-class targets. This result is better than traditional algorithms and is very close to the results of deep-learning methods.

With the development of synthetic aperture radar (SAR) technologies in recent years, the huge amount of remote sensing data bring challenges for high-speed transmission and real-time processing. The general platform with GPU, CPU or digital signal processor (DSP) cannot meet the requirement for short delay and low power. Therefore, a field-programmable gate array-digital signal processor (FPGA-DSP) SAR imaging accelerate platform has become a solution for processing performance with a lower latency and power. In this study, an effective mapping strategy is adopted to the FPGA-DSP hybrid heterogeneous architecture to accelerate the processing performance according to the analysis of Chirp Scaling SAR imaging algorithms. As a proof of concept, a FPGA-DSP hybrid heterogeneous accelerating platform is designed and realised. The platform requires 12 s to focus a stripmap SAR raw data with a granularity of 16,384 × 16,384.

In this study, a new method for sea surface target detection in low signal-to-clutter environment is proposed, which can not only reduce the computation amount but also suppress the sea clutter, static targets and noise. Firstly, the authors use a more general echo signal model, which is more matching for the sea environment. Next, a fast estimation technique with few searching times to reduce the computation of the detection process is put forward. Then, a canceller is designed, which has the merit of suppressing the strong sea clutter, static targets, and noise. Finally, the experiment results demonstrate that the proposed method can achieve high detection probability, fast estimation, and strong suppression ability.

In this study, a fast and accurate method to predict the radar cross-section (RCS) of large-scale and complicated shape targets is proposed based on a high-performance parallel finite difference time-domain (FDTD) numerical method. To this end, several most popular parallel computation methods [including OpenMP, graphics processing unit (GPU), and message-passing interface (MPI)] are discussed first. Based on this discussion, a novel MPI–OpenMP–GPU hybrid parallel computation scheme for FDTD is developed. Moreover, the corresponding load-balance parallel configuration is discussed as well. Since this hybrid parallel scheme combines the merits of existing parallel technologies, the computation performance is remarkably improved. The results show that the computation time of the RCS simulation of a large-scale target can be reduced from 3 days to 0.8 h, that is, ∼98.9% time saving.

Cognitive passive radar (CPR) system is amenable to electromagnetic environment cooperating and optimal strategy making tasks on account of illumination source availability and waveform evaluation. In fact, intelligence will provide performance gain from multi-dimension optimisation. Therefore, we propose a multi-standard multi-scheme compatible wideband CPR framework which selects the most suitable waveform as the passive radar illumination source. First, we develop a software defined radio platform to realise wideband and flexible signal processing, which provides the feasibility of cognition. Then in order to recognise the electromagnetic waveforms, we present a deep learning framework which can be integrated in CPR system as a classifier using convolutional neural networks (CNNs) and long short-term memory (LSTM). Via the proposed innovative passive perception-action cycle, the CPR prototype system demonstrates the real-time, high accuracy modulation recognition ability and the adaptive signal processing ability. Based on our experiments, it can be demonstrated that our CPR prototype system integrated with deep learning is effective and promising.

With the rapid increase of optical remote sensing (RS) images, on-orbit specific target detection has been facing more and more challenges: limited processing environment, wireless bandwidth, space, power and storage. To resolve this issue, visual characteristics in the compressed domain are exploited to represent and recognise specific objects. In this study, a new saliency method and an airfield detection framework in the JPEG2000 compressed domain is proposed. First, the authors exploit the header information in raw JPEG2000 code streams to generate structural saliency map, and a great deal of non-structural regions in RS images are eliminated. Second, airfield candidates are extracted from low frequency wavelet sub-bands using line segment detection and the parallel density model. Finally, the geometrical properties of airfield candidates are represented by the speed-up robust features descriptor, and the support vector machine classifier is utilised to gain the final detection results. The proposed framework is evaluated on self-made dataset. Compared with the existing relevant state-of-the-art approaches, the proposed method reduces the processing time by 60% while guaranteeing the detection rate.

A kind of high speed and parallel hardware architecture is proposed and designed for digital signal processing of high frequency Pulse Doppler radar here. The platform is based on one XC7K410T FPGA, two XC7K325T FPGAs, one TMS320C6678 DSP, and four sets of MT41J256M8 DDR3. The details of implementation including pulse compression, moving target detection and constant-false-alarm rate are described. The simulation results and resource consumption are presented to demonstrate the advantages of the proposed FPGA implementation architecture.

The power amplifier is an important device in the communication system, and its non-linear characteristics will lead to serious distortion of the output signal, reducing the performance of the communication system. In order to solve the problem of non-linearity in a power amplifier, this study proposes an adaptive learning algorithm based on a digital pre-distortion structure, which combines direct learning and indirect learning structure. It improves the accuracy of pre-distorter parameter and has the faster convergence speed. The pre-distortion device parameters are constantly modified to achieve a good linear effect by using the recursive least square adaptive algorithm. The simulation results show that this structure effectively compensates the output signal distortion of the power amplifier, improves the linearisation degree of the power amplifier and reduces the in-band distortion and adjacent channel leakage ratio.

Ship classification in optical images has been challenged by the complexity of various ships, different imaging conditions, and limited labelled images. Traditional methods focus on extracting handcrafted features for classification, but often fails to design well-performed features for complex images. Here, the authors propose a ship classification approach with CNN. It is capable of learning discriminative features itself by supervised learning and achieving good classification performance. They build two small datasets of optical ship images for training and validation, and conduct several experiments. The experimental results indicate that their approach is effective for ship classification.

SAR raw data missing occurs when radar is interrupted for various reasons. In order to deal with this problem, a method is proposed to focus the missing SAR raw data via stage-wise orthogonal matching pursuit (StOMP). A reference function in time domain is designed for the missing raw data. After matched with the reference function, the energy of missing raw data is concentrated in two-dimensional frequency domain. Then, StOMP algorithm is available to recover the matched raw data in two-dimensional frequency domain. The recovered raw data is available to be processed with traditional SAR imaging algorithms. The Omega-K algorithm is chosen to deal with the recovered raw data in experiments. Point target and area target simulations are carried out to validate the effectiveness of the proposed method for azimuth missing SAR raw data. The data missing rate is 50% in both point and area target simulation. The resolution of point targets can reach 0.3 m in both range and azimuth direction.

Hyperspectral image (HSI) classification is a hot topic in remote sensing community; many researchers have made a great deal of effort in this domain. Recently, deep learning-based manner paves a new way to better classification accuracy. However, the flow of information between layers and layers (e.g. max-pooling) in traditional deep architecture turns out to be ineffective. In this study, a novel spectral–spatial classification framework for HSI based on Capsule Network (CapsNet) and dynamic routing algorithm is introduced. The proposed architecture is composed of a hybrid of 1D and 2D convolutional layers and two capsule layers for better and effective mining and combining features. Consequently, experiments on two popular dataset indicate that CapsNet-based framework outperforms traditional CNN-based counterparts. Moreover, this study reveals great potential for CapsNet model in the field of HSI classification.

Small multi-rotor unmanned aerial vehicles (SMR-UAVs) are a promising platform for low-cost synthetic aperture radar (SAR) systems. However, SMR-UAVs usually suffer from serious position errors due to their unstable motion and low actuary position sensors, and autofocus is an indispensable step for their high-quality imaging. An efficient back-projection autofocus method is proposed for SMR-UAV SAR systems by the principle of minimum entropy. The position error estimation model via minimum entropy is derived. The conjugate-gradient method is used to efficiently estimate the position errors. Moreover, to improve the computing efficiency, the strong scatterer areas are estimated as the input of entropy estimation. The effectiveness of the algorithm is demonstrated using both simulation and experimental data.

Detecting targets in maritime domain is difficult due to the complex physical mechanism of the sea clutter. A basis pursuit clutter suppression method based on improved tunable Q-factor wavelet transform (TQWT) is proposed on the analysis of tunable Q-factor. First, the calculation method for selecting an appropriate tunable Q-factor is given. Second, the TQWT with selected Q-factor is adopted to perform the sparse decomposition for the echo signals to obtain the wavelet coefficients. Then, the ‘Basis Pursuit suppressing clutter’ method is used to optimise coefficients. Finally, the target signal after sea clutter suppression is recovered by inverse TQWT. The authors also explored the influence of regularisation parameter on the algorithm performance which provides basis for the practical engineering applications. The experimental results on the CSIR public dataset demonstrated the effectiveness of the algorithm.

Large-scale insect migration may cause severe plant diseases and pests which can exert serious impact on agriculture. Predicting the trajectory and destination of insect migration accurately is an effective way to prevent pests. Here, an insect migration trajectory simulation method was proposed with two stages of insect migration, taking off and cruising, considered. First, four criteria based on actual weather condition were proposed to determine the take-off area of the insect migration. Then, a method based on Lagrange diffusion model was proposed to simulate the cruising trajectory. This step needs to combine with the insect orientation strategy and insect trajectory speed which can be, respectively, provided by entomological radar and weather radar. Finally, by comparing the determined take-off area and the simulated insect cruising trajectory with the monitoring results of the weather radar, the effectiveness of the method was verified.

Here, a simplified algorithm to calculate log-likelihood ratio (LLR) for m-QAM demodulation is proposed. Traditional method calculating LLR for m-QAM demodulation is supposed to do distance computing between any two constellation points so that the high complexity of the hardware implementation is not acceptable. This simplified method is different from the traditional ways that the distance calculation between any two points in a plane can be omitted, indicating that there is an advantage in the amount of computation in this simplified method. Finally, the simplified method is compared with traditional methods from the perspective of complexity and system performance. The simulation results show that the simplified method can greatly reduce the amount of calculation and have little effect on the performance of the system.

Ground clutter is a crucial factor that affects the detection performance of the airborne forward-looking multi-channel linear frequency modulated continuous wave (LFMCW) radar, so the distribution characteristics of studying the ground clutter plays an important role in suppressing ground clutter and improving radar detection performance. In this article, the authors first use the Ward model to divide the effective clutter area into multiple independent clutter patches and calculate the echo signal of each clutter patch. Then the beat signal of each clutter patch is obtained by the de-chirp processing. Next, a Fourier transform is performed on the beat signal of each clutter patch to determine the range ring where the echo of the clutter patch is located, and the peak spectrum signal of the beat signal of the clutter patch in the same range ring is superimposed, thus the authors can realise the simulation of the ground clutter signal. Finally, the simulation results indicate the effectiveness of the proposed method.

Automatic ship detection in SAR imagery has been playing a significant role in the field of marine monitoring but great challenges still exist in real-time application. Despite the exciting progresses made by deep-learning techniques, most detectors failed to yield locations of fairly high quality. Moreover, the ships with variant sizes and aspects are easily omitted especially for small objects under complicated background. To alleviate the above problem, the authors propose an elaborately designed single shot detection framework combined with attention mechanism, which roughly locates the regions of interest via an automatically learned attentional map. This lay the foundation of accurate positioning of extremely small objects since the background interference can be effectively suppressed. Furthermore, a multi-level feature fusion module integrated in top-down and bottom-up manner is adopted to adequately aggregate features from not only adjacent but also distant layers. This strengthens local details and merge strong semantic information, enabling the generation of higher qualified anchors for the efficient detection of multi-scale and multi-orientated objects. Experiments on SAR ship dataset have achieved a promising result, surpassing current state-of-the-art methods.

To achieve verisimilar target deceptive jamming effect, the jamming signals from different targets in different kinds of situation must be acquired. However, it is difficult to be implemented by the methods using deception-template model. In this letter, a novel target deception jamming method against spaceborne synthetic aperture radar based on electromagnetic (EM) simulation is presented. To improve the computational efficiency, an equivalent bistatic scattered fields are used to simulate the scattering characteristics of false target. The proposed method, which consists of the computer-aided-design model, the deceptive jamming model and the involved EM scattered fields, improves the performance of deceptive target in achieving the jamming results. The simulation results demonstrate the validity of the proposed method.

This study researches the potential benefits of clutter suppression in polarisation filtering technology. In the author's laboratory, the antenna of the experimental system has been designed, called dual-polarised Yagi-Uda antenna and the antenna can receive both horizontal and vertical polarisation simultaneously. The experimental results show that the method of improved polarisation filtering can suppress the clutter effectively, which is composed of direct-path signal and multipath clutters.

In order to verify the feasibility of using microwave to study Yellow River ice, the reflected power of pure ice and mixed ice is measured by electromagnetic waves of different polarisations and frequencies in the microwave chamber. The measured data of microwave chamber measurement was processed to obtain radar cross section values. Based on the experiments in the microwave chamber, this study presents a reflection model for calculating the reflection of pure ice. The calculated value obtained by the model is consistent with the measured values, which also proves the validity of the model. The results of this study will provide some basis for the follow-up study of retrieval Yellow River ice thickness.

For moving ship detection in synthetic aperture radar (SAR) images, ship wakes on the sea surface may be detectable, thus leading to interferences in reconnaissance. To address this issue, a wake region detection approach by using multi-feature recombination and area-based morphological analysis is proposed with along-track interferogram-SAR systems. This approach aims to detect the near-ship wake region that is nearby a ship's hull and between two kelvin cusp lines. A novel test is constructed by recombining magnitude, phase, and the local variance of phase in the complex multi-look interferometric SAR image to detect the potential pixels of the ship wakes. Then, with a clustering algorithm based on the proposed test and the pixel spatial distance, the potential wake regions are acquired, and the real ones having sufficiently large areas can be accurately determined. Finally, experimental results on the TerraSAR-X dataset validate the effectiveness of the proposed method.

In through-the-wall radar imaging, compressed sensing techniques have been applicable for small amount of observations to reconstruction walls or targets imaging in the scenes. The authors consider building walls imaging. The proposed approach exploits a low-rank and sparse model to extract the wall returns from radar signals, and then minimises the total variation regularisation problem for reconstructing building walls imaging using the sparsity of their spatial variations. The results of simulation exhibit that the proposed approach effectively enhances the imaging quality of building walls.

Hyperspecral unmixing (HU) is one of the crucial steps of hyperspectral image (HSI) processing. The process of HU can be divided into end-member extraction and abundance estimation. Lots of abundance estimation methods just take some properties of abundance into consideration, such as non-negative, sum-to-one and so on but ignore the noise corruption. However, in practical applications, there are always high-noise bands in HSI due to water absorption, atmospheric transmission, and other inevitable factors, which lead to the estimation accuracy reduction. Here, we propose a new abundance estimation model which takes the mixing pattern of endmembers and low signal-to-noise ratio (SNR) bands of HSI into consideration simultaneously. The constraints considering not only the low-rank feature of abundance but also the sparsity quality of noise are imposed on the new model for more robust results. Adequate experiments both on synthetic and real hyperspectral data have confirmed the superiority of our method.

This study introduces an automatic method for change detection of multi-sensor remote-sensing images (e.g. optical and synthetic aperture radar (SAR) images). As object-based image analysis can effectively reduce the spurious changes and the sensitivity to registration, first, multi-date segmentation is employed to generate homogeneous image objects in spectral, spatial, and temporal, in order to weak the intensity variation effects of multi-sensor images. Then, modified fuzzy c-means (FCM) algorithms are employed to preliminarily classify optical and SAR images, and a criterion is defined using membership values of parcels to select the sample parcels for each class and image. Finally, a change detection principle, which takes statistical properties as the feature space, is introduced to detect changes between multi-sensor images. The experiment results verify that the proposed method is able to cope with optical and SAR images change detection.

Compared with ground radars, space-borne bistatic radar shows great advantages in higher signal-to-noise ratio (SNR) and robuster security. Firstly, this study establishes space-borne geosynchronous satellite-low orbiting satellite (GEO-LEO) space–time two-dimensional model considering the influence of the rotation of the earth together with the curvature of the earth surface. Then the Doppler trajectory distance dependence of the clutter angle is analysed. Finally, the characteristics of the space-borne GEO-LEO model are proved to be beneficial to clutter suppression and ground-moving targets indication. Simulation results show that the space-borne GEO-LEO model has good performance.

The present work develops the mathematical base of ill-posed inverse problems in optics, as well as a method of processing and reconstructing scenes from incomplete information of their Fourier spectra. This problem is a part of the general problem of the eliminating the influence atmospheric and optical distortion when finding and registering an object at large distances under bad atmospheric conditions. The designed programmed methods are also used to process medical images.

The fast Fourier transform FFT processor is an important part of the space real-time signal processing system based on field programmable gate array (FPGA). Since occupying a large amount of logical resources and storage resources, FFT processor is more vulnerable to high-energy particles in space, resulting in single event upset (SEU). This paper presents a novel FPGA scrubbing framework base on dynamic partial reconfiguration technique for a FFT processor to mitigate SEU. The proposed scheme is compared with the blind scrubbing, the reconfiguration time is reduced by 78%. Then, the resource utilisation is 61.5% less than triple modular redundancy scheme. This paper also presents a DPR controller for FFT processor, which is evaluated in terms of hardware resources and reconfiguration time. A comparison to the Xilinx PRC IP shows that multipath delay feedback FFT controller saves 38.6% resources.

As new concept radar combing quantum technology with radar technology, quantum radar is an extension of traditional radar theory and has the potential to break through the limit of conventional radar detection performance. We study quantum radar optimum decision strategies for various signal and background forms and compare them with traditional photon-counting strategies. Quantum optimum strategy gives better detection probability with fixed false-alarm probability, thus reducing needed signal strength to detect the target, but its validity strongly depends on exact knowledge of signal forms. We briefly discuss the influence of deviating signal amplitudes and phases, and find that in the condition that the false-alarm probability is very small quantum optimum strategy is still useful when the signal form varies in a small region.

In this study, a maximum likelihood distance estimation algorithm is proposed for a multi-carrier radar system to estimate the time delay of non-integer sampling intervals. Particularly, the equivalent relationship between time domain delay and frequency domain phase shift and the relationship between target delay and channel impulse response is used. The idea of maximum likelihood estimation is to find the estimation value that maximises the likelihood function. Therefore, the core idea of this algorithm is the peak search. First, channel estimation is performed on the sampled signal in frequency domain to obtain the channel impulse response. Second, phase compensation and peak search is done to estimate the target delay. Finally, for multiple estimation results, the authors statistic the probability density distribution of the target. In addition, they deduce the Cramer–Rao bound (CRB) of the variance of time delay estimation. The measurement accuracy of the algorithm that they proposed approaches the CRB in the high signal-to-noise regime.

This study presents a chirp scaling (CS) phase function generation scheme based on CORDIC algorithm, it adopts single precision floating-point CORDIC processor to implement a variety of non-linear operations which are involved in CS synthetic aperture radar (SAR) imaging algorithm. We extend the range of convergence of CORDIC algorithm and reduce the data width in rotation unit module by adopting a hardware resource reduction scheme. We also adopt a unified CORDIC processor to achieve square root, multiplication, and division operation, which can substitute multiple single-function processors and simplify the complicate arithmetic in CS algorithm. As a proof of concept, we verify the performance of the proposed design scheme on Xilinx XC7VX690T FPGA platform, it is also applied to 16384 × 16384 points target SAR imaging system, the FPGA resource occupancy of the CS phase function generation module demonstrates that the usage of CORDIC processor can effectively save the consumption of hardware resources, and the maximum operating frequency is acceptable. The accuracy of phase functions can satisfy the engineering requirements.

Aiming at the limitation of the coherent accumulation gain of the air defence missile active radar seeker, which is brought by a new generation of manoeuvrable stealth targets, this study proposes a long-time coherent accumulation algorithm based on the blind estimation of relative acceleration between missile and target, which is suitable for the pulse Doppler seeker. By extracting the relative acceleration information carried in the echo signal, the algorithm compensates the original echo signal phase, so that the compensated signal can be refocused in the frequency domain and the seeker detection capability to the large manoeuvring target can be improved. The simulation results show that the seeker sensitivity can be increased by 8 dB when the system false alarm probability is <10⁻⁷ and the intercept probability is >0.99.

The interference of the multipath spreading among the targets and the interior walls will be introduced into the radar image. This study mainly deals with the multipath suppression problem in three-dimensional (3D) through-the-wall radar imaging. The authors propose a correlation-matching multipath suppression algorithm to eliminate the first-order back wall multipath ghosts. Specifically, they first extract the barycentre coordinates of all objects in the original radar image. Second, a decision criterion is provided to measure the relevance between these objects. If two objects are closely related, they assert that they have multipath-target relationship. Finally, they eliminate the multipath ghosts after all objects being determined. At the analysis stage, they assess the effectiveness of the proposed approach through simulation results, which shows its ability of well-removing multipath ghosts.

To solve the problem of the low rate of template matching (TM) jamming decision method under the incomplete jamming rule library condition in traditional electronic warfare, a clustering and resampling – support vector machine (CRS-SVM) jamming decision method, is proposed. For the air-to-air scene of airborne multifunctional fire control radar, the inter-pulse pulse and intra-pulse features are extracted, and a jamming rule library that is divided into the source domain and target domain sample according to the missing rate is constructed in this scene. DBSCAN algorithm is used to cluster the dataset, and the structure information is found. The training samples are generated via resampling. SVM is used to transform the inner product of the target domain space into the inner product of the feature space to establish the feature space hyperplane. The target domain samples are labelled as the corresponding jamming style label to realise the fast and effective decision. The experiment results show that the proposed method can effectively improve the accuracy and robustness of jamming decision compared with the traditional TM method.

Insect migration influences population dynamics, provides ecosystem services, causes heavy crop damages and spreads diseases. The identification of insect migrants is critically important for the study of insect migration. Radar is an effective means of detecting nocturnal insect migrants. The mass, body shape and wingbeat frequency could be measured by the special-purpose entomological radar. These features are potentially valuable for species identification. This paper proposed a method to extract the mass of migratory insects based on ellipsoid scattering model. This model was used to approximate the small insects in the Rayleigh region, and then the inversion formula of the mass was developed based on ellipsoid scattering characteristics. Finally, through analysing statistical data of insect migrants, the formula was modified to fit small migratory insects. The root mean square percent error is 34.92% for estimates of mass. The given method can promote the species identification and biomass quantification of over-flying migratory insects, so as to study the role of insect migration in ecosystems and the pest control.

Scattering entropy (H), scattering angle (α) and anti-entropy (A) are useful parameters in synthetic aperture radar (SAR) image classification. Usually, full-polarimetric SAR data are needed to extract these parameters. In this study, the authors firstly try to predict these parameters from single/dual-polarimetric SAR data using convolutional neural network. Experiments are done on GF-3 polarised SAR database, and promising results are obtained, where the parameters H and α, the average relative error reached is <10%, the parameter A, the average relative error reached is around 25%, and the classification performance based on predictive parameters is around 80%. Furthermore, the predicting performance using different single- and dual-polarisation is compared. The results and conclusions provide a new clue for the applications of single/dual-polarimetric SAR.

A hybrid CS-DL method for aircraft classification in complex electromagnetic environment is introduced. To classify aircraft from interfered radar echoes, the authors propose a novel classification method based on compressed sensing (CS) and deep-learning (DL). After recovering the spectrum polluted by jamming signals by using CS, they exploit sparse auto-encoder (SAE) to extract modulation features and then classify aircraft. The method is tested by 536 flights of three types of airplanes, and the results show that the correct classification rate reaches 75% even when 41% of the pulses are interfered.

Ultra-wide band (UWB) bio-radar attracts lots of researchers, due to its special advantages of non-contact detection. It is widely used in multiple applications such as medical monitoring, rescue and anti-terrorism missions nowadays. However, identifying the target after it was detected and located comes to be a more challenging task. The study aims to investigate the differences of the target characteristics between humans and animals based on ensemble empirical mode decomposition analysis and to test whether these differences can be distinguished. Finally, the authors proposed a new method to distinguish between humans and animals with 500 MHz UWB bio-radar detection system by calculating the energy ratio of micro-vibration signal in human breathing frequency band. Experimental results demonstrate that the energy ratio can be served as a useful parameter, which is promising to distinguish between humans and rabbits.

The interest in microwave imaging system is increasing because it can serve as a useful tool for concealed threat detecting. The current imaging systems have two major challenges: (i) multi-static arrays suffer from complex hardware design and ineffective backprojection-imaging algorithm, and (ii) monostatic linear arrays suffer from large scanning time due to the electronically scanning one by one of the co-located receive–transmit antennas. To address these challenges, this study introduces a novel linear imaging array based on automotive radar chips and sub-array time-division multiple-input multiple-output working principles, which is hardware efficient and compatible fast Fourier transform (FFT) imaging technique. Through simulation, it is shown that the scanning time is only one-fourth of the traditional monostatic linear array and the FFT image reconstruction technique can be used.

Cognitive radar (CR), which uses a closed-loop feedback perception-action cycle system, aims to develop a higher level of intelligence. In this study, the authors develop an efficient Bayesian approach to target detection in CR. The approach, which uses the Neyman-Pearson criterion with priori probability, can improve the target detection effectively. The recursive form of computing priori probability is also discussed. Numerical simulations illustrate that the proposed target detector can improve the performance of detection with the usage of priori probability, which can be computed recursively using the previous measurements.

The deep neural networks has been developed fast and shown great successes in many significant fields, such as smart surveillance, self-driving and face recognition. The detection of the object with multi-scale and multi-aspect-ratio is still the key problem. In this study, the authors propose a bottom-up feature pyramid network, coordinating with multi-scale feature representation and multi-aspect-ratio anchor generation. Firstly, the multi-scale feature representation is formed by a set of fully convolutional layers which is catenated after the backbone network. Secondly, in order to link the multi-scale feature, the deconvolutional layer is involving after each multi-scale feature map. Thirdly, to tackle the problem of adopting object with different aspect ratios, the anchors on each multi-scale feature map are generated by six shapes. The proposed method is evaluated on PASCAL visual object detection dataset and reach the accuracy of 80.5%.

New non-orthogonal multiple access technology is proposed with the development of 5G, which is expected to support higher spectrum efficiency and massive terminal access, it enables the communication system to achieve better bit error rate performance and alleviate the spectrum congestion problem effectively under limited resources. With the design of spectrum sharing for radar and communication getting more attention, 5G non-orthogonal signal has become potential passive radar illumination source. In this article, an orthogonal multiple input multiple output (MIMO) radar waveform design metric by minimising the integration sidelobe level, is introduced firstly. Then based on sparse code multiple access (SCMA) technology, the non-orthogonal waveform is designed as the passive radar illumination sources. The distance and Doppler resolution of non-orthogonal waveform are evaluated by a more generalised MIMO ambiguity function. Simulation results demonstrate that non-orthogonal SCMA signals have competitive performance with orthogonal code division multiple access signals.

The high-precision target echo generation in the field of radar systems is particularly important for large-scale, comprehensive testing of radars. Conventional echo generation techniques often use planar array air feed to simulate echoes. However, it is complicated and expensive to achieve the echo generation with traditional planar array. Therefore, this study proposes a high-precision target generation technique based on one-dimensional linear arrays. The simulation system achieves the target radial high-precision simulation through DRFM technology; the target azimuth high-precision simulation is achieved through differential GPS technology; and the target elevation high-precision simulation is realised through vertical one-dimensional linear array antenna switching. The semi-physical simulation system presented here has the advantages of simple control, low cost, low link loss, easy maintenance, and high simulation target accuracy. Finally, the system was tested in the field to verify the theoretical concept of the article.

Towed decoy can simulate the time domain and frequency domain characteristics of target in radar beam and form effective angle deception. In order to improve the radar ability to countermeasure towed decoy, this study extracts and uses the radar diagonal difference channel in the conventional monopulse radar. The characteristic of diagonal difference channel is analysed. Then, the equations of sum channel, azimuth difference channel, elevation difference channel, and diagonal difference channel are established. By solving the equations, the angles and amplitude of decoy and target are obtained. Decoy recognition method is proposed. The new method can obtain the angles of decoy and target using only one pulse, and recognise decoy steadily using several pulses. The simulation results confirm the effectiveness of the new method.

Synthetic aperture radar (SAR) target feature extraction is a key technology for SAR target recognition. The existing SAR feature extraction methods mainly focus on amplitude information of the SAR image in spite of phase information. This study proposes a new feature extraction method introducing from concept of wavefront in optics domain. The Zernike polynomials are utilised to approximate the SAR echo wavefront, and these coefficients of Zernike polynomials are used as a new target feature for aiding SAR target recognition. Moreover, an experiment is carried out to prove that this method is effective about a target consisted of two scatters with different spacing between them.

This work studies the direct wave suppression technology of passive radar system based on ‘global system for mobile (GSM)’ communication. In order to suppress direct waves, this study proposes an adaptive suppression algorithm based on NLMS criterion. Achieves great suppression of the strong interference signal, and finally realised target positioning and velocity measurement to the actual measured GSM signal.

Multipath interference is the main factor which has influence in super-low-altitude detection. Aiming at solving the problem of detection for super-low-altitude target, anti-multipath methods under concrete environment are studied here. First, antenna radiation pattern of seeker is introduced to modify traditional four-path method. Then, the hybrid method combining physical optics (PO) and method of equivalent currents (MEC) is used to calculate the scattering electric field of target. Compared to the method of moment (MoM), the PO + MEC hybrid methods are of high accuracy. Finally, the multipath interference scattering from interaction between the target and the environment under the condition is studied and analysed in detain. The simulation results are useful in military application.

Radars operating in the HF band have been widely used to detect and track surface ships from their skin echoes; indeed, in many cases this has been the primary mission of the radar. Yet vessels moving on and through the ocean create a family of disturbances collectively referred to as the wake, though a number of different mechanisms and phenomena are involved. It is therefore of interest to explore the possibility of detecting the wake echo and exploiting it as far as possible. In this study, the author presents an analysis of this problem and shows that there is indeed a characteristic signature associated with the Kelvin wake and this contains retrievable information relevant to target discrimination and classification.

The Doppler sensitivity issue is an urgent problem to be solved for the phase-coded radar although it can give the promise of improved range resolution and resistance against various kinds of distortion. Due to the ideal autocorrelation and cross-correlation properties, the complementary sequence is utilised in high-frequency surface wave radar and MIMO radar. However, the Doppler frequency shift affects the autocorrelation property of the complementary sequence. This study aims to solve this problem. Based on the signal model and mathematical derivation, a new algorithm is proposed and the simulation results indicate the effectiveness of this method. This study also puts forward the signal processing diagram and the method to obtain the new code.

Passive radar sparse imaging requires accurate known system observation matrix, but in practical applications there usually exists transmitter and receiver position errors, which would make radar observation matrix partially unknown, resulting in mismatch between echo measurements and observation matrix, thus seriously degrade the performances of traditional sparse imaging algorithms. Here, the passive radar imaging model under position errors of transceivers is established first. Then, an adaptive phase error correction imaging method based on Bayesian compressed sensing is presented, which can deal with transceivers position errors while reconstructing target image simultaneously. Simulation results verify the effectiveness of authors’ method.

To develop a method to forecast soybean leaf damage caused by larvae of the common cutworm, Spodoptera litura (Fabricius), insect dispersion was monitored with an x-band entomological radar and a searchlight trap during the summer soybean season in 2016 and 2017 in western Japan. The insect catch number in the searchlight trap and number of newly damaged leaves in a soybean field were compared to see how much soybean damage the insect's arrival caused. The 2016 results indicated that catch number peaks appeared 3–5 days before the peaks in the number of newly damaged leaves. The time lag corresponded to the egg period. Similar data were found in 2017. However, both the catch number and the number of newly damaged leaves were smaller that year, due to the lower occurrence of the insect. The radar observation showed dispersion at altitudes over the monitoring site when the peaks in insect catch number occurred. The insects often arrived under a northerly wind condition, from the mountainous area to the north. This study suggested that short-term forecasting of soybean leaf damage by radar and searchlight trap may be possible.

The mission geosynchronous – continental land atmosphere sensing system (G-CLASS) is designed to study the diurnal water cycle, using geosynchronous radar. Although the water cycle is vital to human society, processes on timescales less than a day are very poorly observed from space. G-CLASS, using C-band geosynchronous radar, could transform this. Its science objectives address intense storms and high resolution weather prediction, and significant diurnal processes such as snow melt and soil moisture change, with societal impacts including agriculture, water resource management, flooding, and landslides. Secondary objectives relate to ground motion observations for earthquake, volcano, and subsidence monitoring. The orbit chosen for G-CLASS is designed to avoid the geosynchronous protected region and enables integration times of minutes to an hour to achieve resolutions down to ∼20 m. Geosynchronous orbit (GEO) enables high temporal resolution imaging (up to several images per hour), rapid response, and very flexible imaging modes which can provide much improved coverage at low latitudes. The G-CLASS system design is based on a standard small geosynchronous satellite and meets the requirements of ESA's Earth Explorer 10 call.

In order to realise effective target detection with a large baseline radar network interfered by a single radiation source, an adaptive processing method derived from traditional principal component analysis (PCA) and array signal processing theory is proposed in this study. First, the signal model for target detection with a large baseline radar system in an interference environment is established. Second, based on the characteristics of the signal model, an adaptive suppression method composed of high-precision delay alignment and modified PCA is proposed. Finally, the effectiveness of the method in a typical environment is verified by simulation. The results show that this method can accumulate target energy effectively and realise reliable detection in such case, which has great application prospects.

During the past decade, the authors have developed and applied optical remote sensing instrumentation for in situ remote surveillance and quantification of the aerofauna. The sparse structure of aerofauna makes optical focusing challenging, but the authors solved this issue through applying the century old Scheimpflug condition. With this approach, the authors have managed to reduce size, cost and complexity of atmospheric lidars and accomplished an effective tool for ecological entomology capable of counting thousands of insects per hour. Due to the high sensitivity and resolution in time and space, the authors can retrieve target modulation signatures in the kHz range for target classification purposes. As opposed to the cm waves in entomological radar, the authors rely on near infrared (IR) light ∼1 μm. This allows superior beam quality, negligible ground clutter and applications close over ground or within vegetation structure. Near IR light can assess both molecular and microstructural properties of the target through differential absorption and depolarisation. Here the authors give the background of entomological lidar, summarise the authors’ recent progress and put it in context with contemporary work. The authors outline applications, ongoing activities and state of the art. The authors discuss future prospects and challenges.

This study deals with the problem of adaptive detection in the presence of mismatch signals in partially homogeneous environment. To enhance the selectivity of the detector, the authors add a fictitious signal under the null hypothesis. The fictitious signal is orthogonal to the nominal signal steering vector in the quasi-whitened observation space. Two generalised likelihood ratio test-based mismatch selective detectors are derived. Simulation results show that the one-step mismatch selective detector coincides with the adaptive coherence estimator and the two-step mismatch selective detector has a better mismatch rejection capability than the generalised adaptive matched filter.

In the task scheduling problem of phased array radar, the objective function is usually based on the time shifting rate (TSR). However, the TSR cannot achieve expected scheduling performance in non-linear filtering systems. To further improve the scheduling performance, a novel index named shifting improve rate (SIR) is proposed by analysing the influence of time shifting and its formula in Kalman filtering and expand Kalman filtering is derived. Based on SIR, a new objective function for scheduling of phased array radar tasks was designed and used the genetic algorithm to solve it. Simulated experiments are done to validate the proposed index and the new objective function. The results demonstrate that the SIR achieves better scheduling performance and tracking accuracy than TSR in non-linear filtering systems while keeping similar performance in linear filtering systems.

In order to solve the problem that conventional inverse synthetic aperture radar three-dimensional (InISAR 3D) imaging methods need too much computation and so many restrictions in image registration step, this study proposes an InISAR 3D imaging algorithm based on joint translational compensation. The algorithm employs joint translational compensation of range and cross-range direction to achieve the aim of imaging and registration. While obtaining the two-dimensional ISAR images, the accurate registration of different ISAR images is completed. The algorithm needs less calculations and fewer constraints. Simulation results verify the effectiveness of it.

In this work, a novel approach for the simulation of synthetic aperture radar (SAR) images is proposed. Also, this approach aims to interpret the mechanisms of the dominated scattering centres in the SAR image. An attributed model based on typical structures is used to characterise the 3D geometry of the targets. The physical optics and the geometry optics are used to calculate the main scattering mechanisms of the specified typical structures. Also, based on scattering properties and the imaging process, the simulation approach is set up. The experiments based on the Terra-SAR data of an airplane are implemented to test the validity of the simulation. The work presented would be useful for SAR automatic target recognition.

To address radar/ESM track association problem in the presence of sensor biases and different targets reported by different sensors, an anti-bias track association algorithm based on track distance vectors detection is proposed according to the statistical characteristics of Gaussian random vectors. The state estimation decomposition equation is first derived in MPC. The track distance vectors are obtained by the real state cancellation method. Second, in order to eliminate most non-homologous target tracks, the rough association is performed according to the features of the azimuthal rate and inverse-time-to-go (ITG). Finally, the track-to-track association of radar and ESM is extracted based on track distance vectors chi-square distribution. The effectiveness of the proposed algorithm are verified by Monte–Carlo simulation experiments in the presence of sensor biases, targets densities, and detection probabilities.

With the increase of radar signal bandwidth and the application of RF directing sampling technology, wideband radars are facing greater pressures for data transmission, storage, and processing. Based on the relationship between linear frequency modulation (LFM for short) signal and LFM-stepped signal, synthesis method of high-resolution range profile (HRRP for short) for LFM step signal can be applied to LFM signal. Combining the synthesis method of HRRP for LFM signal with compressed sensing theory, this study proposes a new compression method for wideband radar echo with LFM signal style based on random segment selection and analyses the constraint conditions of the proposed data compression method. Finally, the validity and practicability of the proposed data compression method are verified from two aspects: simulation experiment and actual data processing.

Interrupted sampling repeater jamming (ISRJ) uses the under sampling technology to transmit and repeat the radar intercepted signals. As a new kind of coherent interference, it is widely used in self-defence interference. At the same time, the existing radar anti-jamming technology cannot effectively combat it. In this study, the principle of ISRJ is analysed. In view of the mismatch between the jamming signal and the radar transmitting signal, an orthogonal phase-frequency coded signal in a pulse is designed. In signal processing, matching and filtering are processed separately through each sub-signal. The simulation results show that the designed signal has a good inhibitory effect on ISRJ.

In many spatial cooperative vehicle networks, omni-directional communication is preferred due to its flexibility for networking. In the recent studies, space time code was shown to be an effective approach for improving the performance of the omni-directional communication between space vehicles. This paper studies the unified scheme for the orthogonal space time block code–based transmission between two space vehicles equipped with different numbers of antennas. It is obtained that at most six passive antennas and four RF chains are enough for the typical use. In the proposed scheme, the receiver can perform a unified detection procedure regardless of the actual number of the transmit antennas, which facilitates the practical application. Simulation results validate the proposed scheme in terms of the cumulative distribution function of the received signal-to-noise ratios.

With the development of modern electronic warfare, the jammer is becoming smarter than before and can be able to take actions according to the intercepted radar waveform. It has considerable significance for the radar to design a proper waveform to improve its performance when facing the smart jammer. Here, a novel cognitive radar waveform design method is proposed to improve the radar's performance. The relationship between the radar and the jammer is modelled as a two-person zero-sum Stackelberg game, where the cognitive radar is the leader and the jammer is the follower. Signal-to-interference and noise ratio (SINR) criterion is utilised as the utility function, where the radar tries to maximise the SINR while the jammer aims to minimise it. Lagrange multiplier method and particle swarm optimisation (PSO) algorithm is applied to solve the problem and the Stackelberg equilibrium (SE) is obtained. Numerical results show the effectiveness of the proposed method.

Ship detection is an important direction of synthetic aperture radar (SAR) image application in maritime surveillance. A multi-scale optimisation threshold saliency detection is developed in this letter, which is used to detect ships targets of SAR images. The SAR image is first decomposed into a pyramid image sequence. Then the saliency detection is performed using the spectral residual method for each layer in the sequence, and the salient sub-graph that contains ship targets is obtained. Finally, each sub-graph is fused and the optimisation threshold segmentation method that applies to the saliency map gives the final detection result. Experimental results show that the proposed approach has low complexity and high detection accuracy and greatly reduces the dependence on prior knowledge.

This letter addresses the problem of target relocation in wide-area ground moving target indication system. The antenna pattern information is exploited as prior knowledge to enhance the performance of airborne radar. Firstly, the ideal target relocation model for moving targets based on the dual-channel radar system is analysed and the mismatch model of the receiving channel is constructed. Then the reason for poor relocation performance is analysed and the channel error model is constructed. Based on this fact, the problem of target relocation is converted into a problem of channel phase estimation. Moreover, a least-squares method is utilised to estimate the phase error by exploiting the real antenna pattern information and a modified mono-pulse curve (MPC) is derived. Finally, the real geometry location of the moving target can be achieved with the modified MPC. Airborne radar experiments are given to verify the effectiveness of the proposed algorithm.

(Dis)similarity measures play an important role in the interpretation of polarimetric synthetic aperture radar (PolSAR) images. Here, the authors introduce a kind of similarity measures for PolSAR images based on the concepts of Hölder pseudo-divergence and Hölder divergence. Authors’ similarity measures are more generalised as the derived formulas indicate that they contain several widely used measures, such as Bartlett or Bhattahcharyya distances and Chernoff distance. Also, their similarity measures are derived from the complex Wishart distribution, so these measures are good at quantifying the similarity between two covariance matrices and perform well while dealing with classification problems for PolSAR images. Experimental results on unsupervised and supervised classification of PolSAR images also verify the effectiveness of their measures.

In through-the-wall human imaging, ambiguities in wall parameters will cause image blurred and deteriorate signal-to-clutter ratio of vital signs. Autofocusing method is a kind of method to get focusing image without any information of wall. This method is based on the idea that the focusing performance improves when estimated wall parameters are close to the true parameters. In traditional autofocusing methods, the interesting targets such as the point-like targets are selected as the reference targets to assess the focusing performance of image. However, in some cases, especially in through-the-wall imaging of human vital signs, the existence of human cannot be assured before detection and discrimination. Nevertheless, high-performance detection of human is the aim of imaging. The wall always exists in the imaging area, and reflections of wall edges are stronger than the interested vital signs. Thus, in this study the authors select the edges of wall as the prominent targets. The linearity degree of wall edges is adopted as the assessment item of focusing performance since the images of wall edges after compensating the wall parameters are lines approximately. Simulation results show that the proposed autofocusing method can effectively improve the focusing performance of human vital signs.

Falls in the elderly represent a serious challenge for the global population. To address it, monitoring of daily living has been suggested, with radar emerging to be a useful platform for it due to its various benefits with acceptance and privacy. Here, we show results from the use of an S band radar for activity detection and the importance of selecting specific frequency bins to improve its suitability for human movement classification. The use of feature selection to improve detection of key activities such as falls has been presented. Initial results of 65% are improved to 85% and further to 90% with the aforementioned methods.

Sliding spotlight synthetic aperture radar (SAR) is a hot imaging mode in space-borne SAR. It can acquire high-resolution and large azimuth scene size simultaneously. The main challenges for sliding spotlight SAR processing in real time are the huge volume of raw data and complex imaging algorithm. The rapid development of the graphics processing units (GPUs) provides a new computing platform for high-performance SAR processing. This work presents a parallel processing of sliding spotlight SAR imaging algorithm on GPU via compute unified device architecture and tests the speedup of each module of the algorithm on the K20c GPU platform from the aspect of a small granularity and a large granularity of point target images. Compared with traditional CPU-based SAR imaging technology, the proposed GPU-based sliding spotlight SAR imaging obtains a speedup of 186.80 when processing data with a granularity of 16,384 × 8192 and gets a speedup of 49.72 when processing data with a granularity of 65,536 × 32,768, thereby greatly improving the real-time imaging processing.

Inter-modulation would occur when the transmitting system is under multi-tone excitation; therefore, the non-linear response of power amplifier unit should be considered in the system design. Here, by establishing non-linear theory simulation model of the transmitting system, we introduce the channel consistency factor, put forward the concept of effective jamming power, analyse the effective power of the multi-channel array transmitting system under multi-tone excitation, and carry out relevant experiments for verification, thereby providing reference for the design of multi-signal transmitting system.

The computational cost of many deep convolutional neural networks (CNNs) for the automatic target recognition proposed in the area of synthetic aperture radar (SAR) imagery recently is huge, and the limited SAR images are always insufficient for training the deep CNNs. To improve the computational efficiency, a new light but very efficient convolutional network architecture is designed using some novel techniques to get the better results. The authors apply the batch normalisation before each convolutional layer in order to reduce ‘internal covariate shift’ and use the drop-out strategy in the fully layer to avoid over-fitting. Additionally, concatenated ReLU is used as activation scheme specially instead of ReLU for preserving the negative phase information to get the double feature maps of the previous layer rather than to increase the depth of the filters that can lessen the parameters of the networks. The results of the experimental demonstrate that the authors’ CNNs can both achieve a state-of-the-art classification accuracy of 99.53% of the SAR target classification in the Moving and Stationary Target Acquisition and Recognition ten classes public dataset and perform well even when the training data is sparse.

Due to the limitation of laser modulation technology, the azimuth Doppler ambiguity problem exists in the process of inverse synthetic aperture ladar (ISAL) imaging for spinning targets. The traditional azimuth imaging method will not be used to obtain a good two-dimensional image. Therefore, we consider using the target's spinning information for imaging. The spatial geometric model of the spinning target ISAL imaging is established, and the characteristics of the echo signal are analysed. An ISAL imaging algorithm based on the backward projection transform is proposed. First, the spinning angular velocity of the target is obtained by the generalised autocorrelation method, and then the envelope and phase of the distance and the slow time domain are transformed into a backward projection to achieve coherent accumulation, and the two-dimensional high-resolution image of the spinning target is finally obtained. Due to the use of echo phase information, the sidelobe effect is low and the resolution is high. The simulation results show that the algorithm can still get well-focused images under low SNR and Doppler ambiguity.

To improve the RadCom system of performing both data transmission and radar-sensing functions, here a waveform design method for orthogonal frequency division multiplexing (OFDM) signal is proposed to solve the problem of high range sidelobe level in radar detection process while maintaining a reasonable channel capacity. The method only transmits data in the sub-channels with preferable channel quality. Then, calculate the weight of the chosen subcarriers through the convex optimisation method to traditionally minimise the peak-to-sidelobe (PSL) ratio. Simulation results prove that, compared with traditional range sidelobe suppression method, this algorithm is more suitable for RadCom system.

To avoid decline of FDOA estimation precision caused by relative Doppler companding, Doppler rate has to be considered and estimated jointly with TDOA, phase, and FDOA. However, CRLB of the joint estimation is still unclear. This study addresses on derivation of the CRLB. First, the joint estimation model is constructed. Then, the signal-specific CRLB for any type of signal and a simplified CRLB for stationary and constant-envelope signal are derived, respectively. Finally, as a verification, the derived bound is compared with previous bounds for the special cases of TDOA, phase, and FDOA estimation, which shows that authors’ bound is reasonable. In addition to the conventional scenarios of TDOA, phase and FDOA estimation, their bound can also indicate the error characteristic of Doppler rate, which can be used to evaluate estimation performance and forecast location accuracy using Doppler rate.

Using the statistical signal processing principle in the field of natural language processing, a radar state identification approach based on the hidden Markov model (HMM) is proposed. Since each radar state is modelled by three model parameters of a HMM, the radar state identification can be solved from the solution of the evaluation problem of a HMM. Simulation results show that the HMM-based statistical identification method has tolerance to parameter error, which is suitable for the intelligent identification of the radar state in a complex environment.

Entomological radars need to address the problem of insect density estimation. Here, a Ku-band experimental entomological radar with two-dimensional scanning mode is used to collect insect echo data. The nearest neighbour method is used to associate the traces of the same insect. Then, the insects are classified according to their radar cross-sections (RCSs) and heights. Then, the corresponding radar-effective sampling volumes are calculated. So, we can get the density of insects with different RCSs at any height layer. Finally, the densities of insects with different RCSs at the same height layer are accumulated to obtain the total insect density at each height layer.

Cloud detection plays a significant role in remote sensing (RS) image processing. Numbers of cloud detection algorithms have been developed in the literature. However, they suffer the weakness of omitting thin and small cloud, and poor ability of differentiating the cloud from confusing ground region (e.g. artificial building). In this study, a robust ragged cloud detection algorithm for RS image is proposed. First, the simple linear iterative clustering method is applied to segment ragged cloud. Then, the improved Qtsu's method is introduced to remove the redundant superpixel. Finally, the Natural Scene Statistic is designed to classify the cloud region. Finally, original image will be classified into thick cloud, thin cloud and non-cloud. Experimental results indicate that the proposed model outperforms the state-of-the-art methods for cloud detection.

The range profile of the plasma sheath target is always found to be distorted when using the linear frequency modulation signal as the radar detection signal. As a result, the detection and tracking of the target became hard, or even failure. Here, three possible reasons of the range profile distortion is investigated in detail: (i) The broadening of the particles in the radial direction of the radar detection when the plasma sheath is produced; (ii) The scatterers in one range gate have different radial velocities that the range profile is dispersion due to the range-Doppler coupling effect; (iii) The amplitude and frequency of the echo signal are modulated by the time-varying plasma sheath. Here, the influence of the above three mechanisms on the range profile is simulated. In addition, for the accurate detection and tracking of these targets, the reason of the distortion should be distinguished. Therefore, several methods are proposed to distinguish the different mechanisms for the range profile distortion.

The backscattering of radar targets is sensitive to the relative geometry between targets' orientations and the radar. This target scattering diversity makes mechanism interpretation and quantitative application difficult. This study reviews a newly established theory for target scattering mechanism interpretation in the rotation domain. It contains uniform polarimetric matrix rotation theory and the interpretation tool of polarimetric coherence pattern. The core idea of the proposed method is to expand targets' information obtained at a specific orientation to the full rotation domain, which can provide fundamentals for the deep mining and utilisation of target scattering information. On this basis, typical applications with the derived new polarimetric features are introduced and the effectiveness of these features is demonstrated.

This study develop an adaptive stationary target imaging technique based on the empirical mode decomposition (EMD) method for the time-division multiple-input multiple-output through-wall radar. By exploiting the discrimination between the respiratory rate and the environmental clutters' frequency, the weak echoes of the stationary target are enhanced via EMD processing adaptively. Based on the simulations as well as the experimental data, the authors verify that the stationary target can be positioned accurately via their proposed method.

The target angular glint is the main factor restricting the tracking accuracy of radar at close range. There are few researches on angular glint of pulse compression radar. Here, the angular glint model of the large aerodynamic target under pulse compression is proposed. The field measured data verify the validity of the model. Second, a frequency-diversity algorithm based on gating weighting is proposed to suppress angular glint. Simulation results show that its performance is better than the traditional frequency diversity algorithm. Third, the influence of total frequency hopping bandwidth and number of hopping on the frequency diversity algorithm is discussed. Finally, the algorithm proposed here has been applied with the existing experimental prototype. The outfield experiments show that the algorithm can suppress angle glint noise effectively.

Ship detection from space-borne SAR image is a challenging task. The main challenges in this task include the disturbance of sea clutter, island, coastline, and variability in ship sizes. Here, the authors propose a novel hierarchical ship detection framework to overcome the above challenges. Initially, the edge-enhanced maximally stable extremal regions (E-MSER) is employed as ship candidates. Then, through some simple shape analysis, they can eliminate the obvious false alarms which are not conformed to the shape feature of ship. Finally, candidates are represented by an effective multiple features learning framework for discriminating true ship targets. Experimental results on numerous space-borne SAR images demonstrate their proposed scheme that outperforms other traditional methods with higher detection accuracy and lower false alarm.

The sequential Monte Carlo cardinality balanced multi-Bernoulli (SMC-CBMeMBer) filter provides a good framework to cope with the multi-target tracking problem. However, the standard SMC-CBMeMBer filter suffers from the particles’ degradation problem seriously. Using the measurements to construct the proposal density in the step of predict can effectively solve the above problem, but this kind of approach brings an amount of computation and causes the overestimation of the target number. To examine the quality of each predicted particle adaptively and use the cubature Kalman filter (CKF) to refine the poor-quality particles with the aid of the current measurements is proposed in this study. This method manages to alleviate the particles degradation problem without increasing the computational complexity seriously since only a part of the particles is refined by the CKF. Also, the proposed method can avoid cardinality overestimation caused by abuse of measurements. A range of simulations is performed to test the performance of the proposed method. The results confirm the effectiveness and robustness of the novel method.

GM_PHD (Gaussian mixture of probability hypothesis density) cannot completely track multiple targets, such as the flying birds in the complex low-altitude airspace near the airport, due to the lack of the steps of birth detection, track extraction and death detection. A new algorithm is proposed to solve this problem, which mainly contributes to the following three aspects. First, the k-nearest neighbour algorithm is used to detect the birth of bird targets from measurements which is necessary to construct the birth intensity function. Second, the clustering algorithm is introduced into the probability hypothesis density filter framework to extract the bird targets’ tracks from the filtering results. Third, an algorithm is added to detect the death of bird targets for better tracking. The Gaussian mixture implementation of the algorithm denoted as BT_GM_PHD (Bird Tracking GM_PHD) is presented. The test results on simulation and ground-truth data show that the proposed BT_GM_PHD algorithm can effectively track the multiple flying bird targets in the complex low-altitude airspace near the airport, outperforming the GM_PHD filter.

With the increase of target mobility and environment complexity, highly reliable single-period statistical judgments become more and more difficult. The multiple hypothesis tracking (MHT) algorithm is a method based on delay logic, and can effectively solve the problem of data association in the course of tracking. However, the number of hypotheses generated by the MHT algorithm is exponentially related to the false alarm rate and the number of targets. Therefore, the hypotheses reduction techniques are required for the implementation of the algorithm. The N-scan-back method and K-best method are often used. On the basis of understanding the K-Best method and the N-Scan-back method, this study proposes the two-dimensional constraints by the K-Best method and the N-Scan-back method, and besides adds the target manoeuvre detection method to jointly manage the hypotheses based on the mean of the filter residual. According to the manoeuvre detection results, the optimised MHT algorithm can adjust the likelihood probability calculation model of the plot-track association and the posterior probability calculation model of the hypothetical branch track, and shorten the decision depth N in the N-Scan-back method. Through simulation, it is proved that the optimised algorithm can reduce the calculation amount and improve the tracking accuracy.

Here, the authors propose a robust adaptive beamforming method based on desired signal steering vector estimation and interference-plus-noise covariance matrix reconstruction, so as to attenuate the influences of the desired signal steering vector mismatch and the limited training snapshots on the performance of adaptive beamformer. More precisely, the desired signal steering vector is estimated by minimising the sine value of the angle between the presumed desired signal steering vector and the eigenvectors of sample covariance matrix. Besides, the sample covariance matrix is reconstructed by reducing the dispersion extent of the noise eigenvalues and eliminating the desired signal component from the sample covariance matrix. The proposed method can not only accelerate the convergence speed of adaptive beamforming algorithm, but also avoid the desired signal cancellation phenomenon when the desired signal is present in the training snapshots. Simulation results demonstrate the superiority of the proposed method.

In this study, a total least square (TLS) based cancellation method is proposed for suppressing direct and multipath clutter signals in passive radar. The TLS algorithm is first used to estimate the distribution of the clutter signals over delay bins. Then the clutter signal subspace is constructed by delaying the reference signal according to the estimated number and delay bins of the clutter signals. Finally, the surveillance signal is projected into the orthogonal subspace of the clutter space. Compared to the typical extended cancellation algorithm, the proposed method uses a smaller but more accurate order of clutter space, so it can get accurate estimate of the clutter amplitudes by only using a limited length of signals. This indicates the proposed method is with the potential to suppress time-variant interference where only a small number of signal samples is available.

This study considers the problem of deceptive multiple false targets jamming recognition. Due to the different modulation patterns, there is a great difference in fast-time frequency spectrum or slow-time Doppler spectrum between the different types of jammings. The proposed algorithm first extracts the received sample matrix, then computes left and right similarity (LR-similarity) by fast-time-slicing or slow-time-slicing process and finally recognises the samples according to template matching classifier. Experimental results demonstrate that the accuracy recognition probability of the proposed method is >95%, when jamming-to-signal ratio is above 3 dB.

Radar application in modern warfare becomes more and more rigorous because of the rapidly developed radar countermeasures, especially active jamming in recent years. It costs a radar many resources for anti-jamming in order to detect a target. Hence, it is of great value to recognise the active jamming and thereafter take measures to distinguish target from the numerous jamming. Traditional methods of recognition jamming are blamed for its low efficiency and low accuracy. Radar researchers are looking forward to a new way to do the recognition work. Machine learning has made great advancements in many areas such as image classification, language translation, signal processing and many other recognition tasks, due to its great performance and high accuracy. The authors applied a machine learning method, i.e. convolutional neural networks, to recognise active jamming here. The authors’ results demonstrate that convolutional neural networks have strong ability to distinguish active jamming and thus provide them adequate preparation for anti-jamming process.

Automatic detection of specific artificial region is one of the research hotspots in remote sensing image. In this study, a hierarchical detection method of specific artificial region using local structure constraint in remote sensing images is proposed. The process can be divided into two main stages: screening of keypoints based on the description of structural pattern, and specific artificial regions detection based on the local structural similarity. This method can greatly reduce the calculation of feature description in the redundant non-specific regions, and effectively reduce the error matching points to achieve the rapid and reliable detection of the specific artificial region. The authors test the proposed method on the artificial region sets with different types and time phase. The experimental results show that the method has high computational efficiency and detection accuracy.

Most existing algorithms in high-resolution range profile recognition focus on the closed set cases, where the test sample is from a known class. However, a sample could be drawn from unknown classes in realistic scenario, which is named as open set recognition. Here, open set HRRP recognition is achieved by incorporating extreme value theory into convolutional neural network. The softmax layer is replaced by a so-called openmax layer which estimates probabilities of the test sample belonging to known and unknown classes. Experimental results demonstrate that the proposed method outperforms the state-of-art algorithms such as NN, 1-vs-set machine, and W-SVM in terms of correct rejection rate.

To solve the problem caused by angular glint in the terminal guidance process, which effects steady target tracking and results in the final miss, the influence of angular glint on the seeker's angle tracking and angular glint compensation methodology using multiple output angle error equation with different detection frequencies at single sensing period to extract the unknown parameters of the target model based on the orthogonal multicarrier technique is proposed here. The method is verified reasonable and feasible through the simulation analysis, the analysis result can be applied to improve the tracking precision of the seeker's terminal guidance process.

In this study, a single channel pipelined variable-length fast Fourier transformation (FFT) processor design is presented. With the development of science and technology, digital signal processing has been widely used in many fields. The most basic algorithm of digital signal processing theory, FFT, is widely used in radar signal processing, remote sensing, image processing, communications and other fields. A high real-time and high-throughput FFT processor design method based on the field-programmable gate array (FPGA) platforms is proposed to meet the needs of the synthetic aperture radar imaging digital signal processing. The authors conducted the ISE synthesis based on the Xilinx FPGA platform using xc6vlx760ff1760-1 as the target chip and the ModelSim simulation, combined with Matlab, to verify the correctness of this design.

Here, a wideband signal direction-of-arrival (DOA) estimation method was proposed. By applying the keystone transform, the phase difference caused by different frequency is eliminated, and then the narrowband DOA estimation method can be used directly. Compared with the traditional method, the proposed approach can achieve better estimation accuracy than incoherent subspace method and coherent subspace method at low signal-to-noise ratio. The theoretical derivation of the method is given. Monte–Carlo simulations are provided to demonstrate the performance of the proposed scheme.

In this study, a target classification method is proposed based on a third-order cyclic statistics technique. The authors introduce cyclic bispectrum (CBS) to reveal the non-linear cyclic nature contained by the micro-Doppler signal, and it is observed that the non-zero peaks generated by some cyclic non-linear nature form unique distribution patterns on CBS slices for different targets. Then, a Renyi entropy is calculated for each CBS slice to measure the information content and thus achieve an entropy sequence. Subsequently, considering the entropy sequence as a feature vector, the support vector machine classifier is used to perform the target classification. Experimental results based on real measured data validate the effectiveness of the method.

In the multi-radar tracking system, multi-radar data fusion technology can improve the detection and tracking capabilities of radar. Besides, the spatial registration processing of radar data can effectively calibrate the radar system error and improve the fusion accuracy. In this study, the system error registration for two-station coast radars based on non-cooperative targets is implemented by using the exact maximum likelihood spatial registration algorithm (EML). On the basis of the EML algorithm, this study proposes a selection method for the corresponding targets. The selection method that facilitates the application of the EML algorithm can automatically complete the clipping, alignment and registration of the track information. For the defect that the EML algorithm is easy to fall into the local minimum in the iterative processing, the data inspection method based on the minimum root mean square test is proposed in this study, which can further improve the exactitude of the system error estimation. Compared with direct fusion, the plot fusion accuracy of the two-station coast radars is significantly improved after the spatial registration of the EML algorithm.

In practice, medium pulse repetition frequency (PRF) radar system is widely used as it takes both the advantages of low and high PRF radars. However, range and Doppler ambiguities appear in such a complicated system. Traditional target tracking algorithms first threshold raw measurements and obtain points from different PRF, then perform ambiguity resolutions over them to obtain the unambiguous measurements. It leads to a rather poor performance when target signal-to-noise ratio is low. While multi-frame track-before-detect (MF-TBD) is an advantageous method to track weak targets, but it fails to be applied in multiple PRF radar system directly. In this study, two algorithms based on MF-TBD are proposed. Classical ambiguity resolution (CAR) TBD first resolves ambiguities by means of a robust Chinese Remainder Theorem, where the first detection threshold is below the normal threshold, and then MF-TBD is applied for further tracking. The simulation shows that the proposed method significantly outperforms the CAR detect-before-track algorithm.

In modern electronic warfare, integration of radar and communication plays an important role in maximising resource utilisation and reducing electromagnetic interference, especially for the integrated waveform design technology based on signal sharing. Here, the authors introduce polyphase complete complementary codes (CCC) into the integration of spread spectrum radar and communication. According to the correlation property of CCC, they deduce the analytical expression of the ambiguity function of CCC and introduce the integration system based on CCC. Simulation results show the superiority of CCC in correlation, which means that it can be used as a spread spectrum sequence in the radar and communication integration field.

A novel coherent integration method for high dynamics direct sequences spread spectrum (DSSS) signal is presented. The presented method is based on adjacent cross-correlation function (ACCF) and Lv's distribution (LVD). ACCF is firstly used to compensate the code phase drift and reduce the Doppler frequency migration order. After that, LVD is used to accomplish the integration and obtain the parameter estimates. The presented method can estimate motion parameters without any searching procedure. Compared with the existing methods, the presented method can achieve a good balance between the computational cost and the detection and estimation performance. Numerical simulations are provided to verify the effectiveness of the proposed method.

Most of the signals received in the field of wireless communication, radar, and sonar are wideband, so there is an active research on wideband direction-of-arrival (DOA). Many approaches of DOA estimation for different frequency with same angles have already been proposed, such as off-grid sparse learning via iterative minimisation (OGSLIM). Based on the OGSLIM algorithm, this study proposed frequency group decision strategy (FGDS). Based on the different information of the number of signals at different frequency bins, FGDS can obtain the estimation results of DOA at each frequency bin. The results of simulations demonstrate that the proposed algorithm outperforms other reported ones.

Serious security risks and economic losses have increased rapidly under the threat of foreign object debris. Most detection methods employing optical devices are limited in severe weather for low visibility. This study proposes an applicable approach for moving foreign object debris on the basis of space-time adaptive processing. The geometrical model of an airport is introduced whereby the runway and lawn are divided into clutter cells. After calculating the complex amplitude of clutter in each cell, the authors employ the minimum variance power spectrum and degrees of freedom to describe the ground clutter characteristics. Then the space-time filter is designed under the linearly constrained minimum variance criterion to suppress the clutter and detects the moving target. Simulations show that the errors in azimuth and Doppler are both less than −15 dB when SCNR = −20 dB.

This study proposes a multi-ary modulation chirp spread spectrum algorithm specific for the unmanned aerial vehicles field. It further introduces the corresponding fractional Fourier transform de-spreading algorithm and the hardware implementation of the software radio-based communication system to validate the algorithm's better anti-interference performance in the wide-band non-stationary interference environment.

As Moore's law meets bottlenecks, the demand for heterogeneous parallel processing systems is increasing. Field-programmable gate arrays (FPGAs) are becoming more efficient acceleration devices due to their powerful processing performance, and the CPU + FPGA architecture under the OpenCL framework has become the trend of heterogeneous parallel processing systems. This study focuses on the optimisation of pulse compression algorithm in FPGA based on OpenCL, which plays an important role in modern radar signal processing systems. By using double cache for ping–pang storage of data between matched filter and inverse fast Fourier transform (IFFT), an optimised processing method is proposed by using a pipeline and verify the method by using Arria 10 GX1150 FPGA with two groups of 2 GB DDR3; the results show that the proposed method can achieve 2.89× performance improvement over the conventional implementation.

The estimation of target moving trajectory by the multiple receivers based on global positioning system (GPS) forward scattering radar has been first presented in the study, which cannot be obtained by the traditional method. A new estimation approach has been considered using the following information. First, the starting and ending time at which the target pass through the baseline of GPS transmitter–receiver can be obtained. Second, the azimuth and elevation angles of GPS satellites are extracted in combination with the satellite sky map at the recording time. On the basis, the target moving trajectory can be obtained by the established model. In order to validate its feasibility, the experiment has been completed in the open area of Beihang University, which regards the human as the target. Four GPS receivers have been employed to record the data. The experimental results show that the estimation of target trajectory is consistent with the real situation, which proves the practical applicability of the proposed technique.

Aimed at the problem that the unmanned aerial vehicle (UAV) target is difficult to be detected in the meteorological clutter because of their close velocity, a new detection method is proposed for the ball-borne radar based on Kalmus filter. Firstly, the radar echo is compressed by pulse compression. Then, the meteorological clutter is filtered out by the velocity-corrected Kalmus filter. Finally, CFAR is employed for detection. The simulation results show that this method can effectively suppress the meteorological clutter and improve the detection performance of the ball-borne radar to UAV target.

In this study, ground target recognition based on one-dimensional convolutional neural network (CNN) is studied by exploiting the targets’ high-resolution range profiles (HRRPs). Contrary to conventional methods which need feature extraction artificially, CNN can automatically discover features for classification. The authors propose a multi-channel CNN architecture that can be applied on diverse forms of HRRP such as amplitude, complex, spectrum etc. Experimental results demonstrate the superiorities of the proposed method over conventional methods based on handcrafted features and single-channel CNN in terms of recognition accuracy. Visualisation of the ‘deep features’ shows higher separability than handcrafted features, thus providing an insight into its effectiveness in exploiting the intrinsic structures.

The direction of arrival (DOA) estimation for off-grid targets in the coprime array, which can extend degrees of freedom, is studied. Among existing research results, joint orthogonal matching pursuit (JOMP) has obvious advantages in complexity, but drawbacks in accuracy. Therefore, a bi-level iterative algorithm is proposed to improve estimation performance of JOMP in this study. First, JOMP is executed to obtain the initial estimations of DOAs and mismatch parameters. Then, an iterative proximity searching is implemented to update target vectors to decline the residual further. After that, on the strength of results of the secondary iteration, the signal spectrum and mismatch parameters can be obtained through exploiting LS algorithm. Simulation results verify effectiveness of the proposed algorithm.

Matrix transposition is a very critical operation in synthetic aperture radar (SAR) imaging systems. This study presents an improved corner turning memory solution for real-time SAR imaging processing. Based on the sub-matrix three-dimensional mapping method, the efficiency of matrix transposing is greatly increased. However, in sliding spotlight mode, a novel SAR imaging mode, the amount of raw data is increased too large to directly use the above method. In this study, a joint address mapping strategy for two pieces of double data rate 3 synchronous dynamic random-access memory is used to meet the needs of the amount of data and make up for the lack of original method. The experimental results show that this method can get an improved and balanced two-dimensional access efficiency for sliding spotlight mode SAR imaging processing.

In the radar systems, such as frequency source, power amplifier, mixer, filter and A/D converter, ageing devices, temperature and environmental changes etc., it is inevitable to introduce wideband amplitude/phase error. These can cause seriously channel mismatch that will affect the performance of digital beam-forming, synthetic aperture radar image, space time adaptive processing (STAP), automatic target recognition (ATR), target detection, clutter/interference suppression, pulse compression results and cancellation ratio etc. So this study mainly discusses that the channel amplitude/phase inconsistency of wideband digital array radar affects the pulse compression, beam forming and so forth. It also researches and analyses channel equalisation method, and the simulation and experimental results based on the experimental data show that the method based on Fourier transform is effectively and convenient. The Fourier transform method is easier for hardware implementation that can be applied to practical engineering.

As the traditional imaging dead zone, how to achieve high angular resolution in forward-looking region is always a big challenge in radar imaging field. In recent years, many approaches are proposed to solve this problem. However, most of the proposed forward-looking imaging approaches fail to work in the complex sea-surface situation. In this study, a Bayesian deconvolution method aimed at the sea-surface target is presented, which relies on the convolution model between the target distribution and the antenna pattern of scanning radar. In view of the situation of sea surface, Rayleigh distribution is adopted to describe the statistical property of sea-surface clutter, and lognormal distribution is employed to represent the prior distribution of sea-surface target. Then, the optimisation theory is utilised to obtain the iterative estimated solution. The simulation results are given to verify the superior performance of the presented method.

In this study, the problems, which simultaneously extract and direct positioning of multiple transmitters, are considered. Without assuming a priori knowledge of the source number, the multiple signal classification (MUSIC) algorithm is applied after using the Akaike information theoretic criteria for model order determination, while the minimum variance distortionless response (MVDR) algorithm can be directly used to localise multiple sources that transmit unknown signals. Two methods, which combine MUSIC and MVDR with the image expansion (IE) algorithm, are proposed. The IE algorithm used in actual scenarios is based on the constant-false-alarm rate. Simulation results show that the proposed algorithms can effectively extract and accurately localise multiple transmitters.

In this study, the problem of low-complexity direction-of-arrival (DOA) estimation is addressed, and a novel real-valued propagator method (PM) is presented with a uniform linear array. The covariance matrix is divided into two subarrays and an equivalent noise subspace is obtained by exploiting the standard PM algorithm without eigenvalue decomposition. By a coordinate mapping technique, the complex PM cost function has been converted into a real-valued polynomial whose order only rely on the number of arrays. Using such a mathematical fact, source DOAs can be estimated by polynomial rooting instead of peak searching. The proposed method is able to reduce significant complexity with comparable root-mean-square error performance to the standard PM, which is finally verified by numerical simulations.

Most existing velocity measurement methods can only calculate the radial velocity by Doppler effect. Here, the authors study the method of measuring the lateral velocity. It is found that the lateral velocity influences the changing rate of Doppler frequency over time, which inspires them to measure the lateral velocity by analysing the Doppler spectrum or the quadratic phase of the echo. The relationship between the echoes and the lateral velocity is modelled in this study. Also the mean square error of lateral velocity measurement, including the model error and the estimation error, is also deduced. The influence of the observation time is studied according to the mean square error.

This study presents a novel super-resolution directions of arrival (DOA) estimation method for mechanical scanning radar by using the advanced compressive sensing algorithm. This method is implemented by constructing a signal model of the mechanical scanning radar through imitating the array radar signal model. Also then it established the compression relationship between the transmitted and received echo signals of the reader by sparsification technique of sending signals. Finally, the DOA estimation can be solved by employing advanced compressive sensing algorithm including basic pursuit (BP), orthogonal matching pursuit (OMP), regularised orthogonal matching pursuit algorithms and so on. Compared with the conventional super-resolution DOA estimation method such as multiple signal classification, the proposed method can distinguish two or more targets within one beamwidth more precisely. Compared with the earlier super-resolution DOA estimation method such as maximum likelihood, the proposed method can obtain more accurate and robust angle difference estimation and do not need to previously know the source number. Both simulated and experimental results show that the overall performance of the proposed DOA estimation is better than that of other methods.