Search

Highlights • The reference signal reconstruction improves the detection probability. • The efficiency of static-clutter suppression is enhanced by the reconstruction. • The optimum DVB-T signal reconstruction outperforms the conventional approach. • The reconstruction enables exploiting distant transmitters for detection. Abstract The present work investigates the reference signal reconstruction method for passive coherent location (PCL) radars exploiting Digital Video Broadcasting-Terrestrial (DVB-T) signals. The reference signal reconstruction improves the signal-to-noise ratio (SNR) of the reference signal, which enhances the detection probability for DVB-T based PCL systems. The existing approach for the reference signal reconstruction is performed by demodulating the received reference signal to retrieve the transmitted quadrature amplitude modulation (QAM) symbols and modulate them to obtain the reconstructed signal. For low SNR values, the QAM symbol detection is accompanied with a significant error, which leads to a mismatch between the reconstrued signal and the exact one. Consequently, the expected improvement due to the signal reconstruction is limited. In this work, we propose an optimum reconstruction method which filters the detected QAM symbol prior to the remodulation step. To validate the proposed method, we retrieved analytic expressions for the detection and false-alarm probabilities and verified them through Monte-Carlo simulations. In addition, real-data sets were used to evaluate the proposed method performances. The results show that the proposed method outperforms the conventional one by extending the feasibility of the reference signal reconstruction for low SNR values. [ABSTRACT FROM AUTHOR]

Highlights • A new two channel Convolutional Neural Network combining with Bi-directional Long Short-Term Memory achieves excellent performance for modulation recognition and protocol recognition. • Transfer learning is firstly introduced in waveform recognition filed to solve the model transferability problem across different recognition tasks, which can efficiently save the 60% training time and 50% training data. • A complete protocol signal dataset is provided for public research about waveform recognition in passive radar. Graphical abstract Abstract Passive radar capable of recognizing illumination of opportunities can improve the detection performance on account of its functional properties of environment adaptivity. Waveform recognition approaches based on Deep Learning can outperform traditional methods based on hand-crafted feature as shown in recent studies. In this paper, we propose a novel transferred deep learning waveform recognition method which makes use of multi-scale convolution and temporal dependency characteristics to improve the recognition performance. Firstly, we develop a two-channel convolutional neural networks combining with Bi-directional Long Short-Term Memory (TCNN-BL) architecture to extract features of different scales and merge past and future states. Then in order to solve the transferability problem across various sampling frequencies, we present a parameter transfer approach which initializes target domain classifier using source domain parameters. Based on our experiments on both public datasets and our own datasets, it can be demonstrated that the proposed approach significantly outperforms the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

This paper investigates the problem of locating a moving target using a passive radar system with multiple transmitters and multiple receivers. The bistatic range and bistatic range rate between each transmitter and receiver are used as the measurements. A novel algebraic solution employing two-step weighted least squares (2WLS) minimizations is proposed. In the first step, the measurement equations are linearized by introducing multiple additional parameters and a WLS minimization is used to obtain a rough estimate; then in the second step, the known relation between the additional parameters and the target location parameters is utilized to refine the estimate. Theoretical accuracy analysis indicates that the proposed algorithm achieves the Cramer-Row lower bound, and Monte-Carlo simulations demonstrate that the proposed algorithm outperforms existing algorithms. [ABSTRACT FROM AUTHOR]

The target detection problem in multi-static passive radar systems (MS-PRS) is investigated, where multiple transmitters operate at a single frequency and a single multi-antenna radar receiver processes the received signals. In contrast to multi-frequency MS-PRS, the detector design is challenging due to the fact that the detector sees a mixture of transmitted signals which may not be separately treated. We propose the design of generalized likelihood ratio test (GLRT) receivers to optimize the detection performance. The beamformers at the radar receiver are optimized so that the received signals can be seen as statistically independent signals from multiple reference channels and a surveillance channel. The detector optimization problem is accurately approximated with a convex optimization algorithm. It is analytically shown that the performance of the proposed GLRT receiver approaches that of the active radar. Computer simulations verify the analytical results. [ABSTRACT FROM AUTHOR]

This paper examines the target detection problem for a passive multistatic radar employing illuminators of opportunity (IOs), where the receivers are contaminated by non-negligible noise and direct-path interference (DPI). A parametric approach is proposed by modeling the unknown signal transmitted from the IO as an auto-regressive (AR) process whose temporal correlation is jointly estimated and exploited for passive detection. The proposed solution is developed based on the generalized likelihood ratio test principle, which involves non-linear estimation that is solved by using the expectation-maximization (EM) algorithm. We also discuss the initialization of the EM algorithm and the joint adaptive model order estimation for the AR process without using any training signal. In addition, we extend several conventional passive detectors, which were introduced by assuming no DPI is present, to provide them with an ability to handle the DPI problem. A clairvoyant matched filtering (MF) detector is derived as well assuming the knowledge of the IO waveform. Extensive simulation results are presented, using simulated waveforms whose temporal correlation can be easily controlled, as well as practical IO waveforms transmitted by frequency modulation (FM) radio. The results show that the proposed EM-based passive detector outperforms conventional passive detectors due to the exploitation of the waveform correlation. [ABSTRACT FROM AUTHOR]

We propose a method to detect a target in a bistatic passive polarimetric radar network, with weather surveillance radar as our illuminator of opportunity (IO). We build our signal model using electromagnetic vector sensors (EMVS) as the receiver, which captures the reflections from a point-like target present in the scene of interest, surrounded with strong clutter. We develop a generalized likelihood ratio test (GLRT) detector that is robust to inhomogeneous clutter. We also develop a maximum likelihood (ML) solution to extract the signal subspace from the received data contaminated by the clutter interference. We provide the exact distribution of the test statistic for the asymptotic case and evaluate its performance loss by considering a reduced set of data. The proposed GLRT method is a constant false alarm rate (CFAR) detector, which makes it robust against the inhomogeneous clutter. With the help of numerical results, we demonstrate the robustness and the limitations of our proposed method. [ABSTRACT FROM AUTHOR]

• A random measurement matrix is designed to jointly obtain compressive observations and remove the DPI and clutter. • An order-statistic based detector, referred to as the OSOMP, is proposed to address the case where the target support is unknown, i.e. the locations of the IOs/receiver are unknown. • Since OMP applies an iterative ranking process to select the components/atoms of the dictionary, the OSOMP test variable is an order statistic, which has a non-convergent distribution. To cope with this problem, a modified test statistic for the OSOMP detector is presented and an analytical expression for the probability of false alarm is obtained. • To further reduce the amount of calculation, we discuss the minimum number of iterations required by the OSOMP algorithm to achieve the desired probability of detection and false alarm. All MC simulation results show that in the case of unknown target support, the OSOMP detector outperforms the ED in low to moderate SNR regions. This paper considers the problem of compressive target detection with direct-path interference (DPI) and clutter in a single-frequency network (SFN) based multistatic passive radar system (MS-PRS). Specifically, a measurement matrix is designed to jointly obtain compressive observations and remove the DPI and clutter. We first analyze a compressive subspace detector which assumes the target support is known. When the target supports cannot be accurately obtained, an order-statistic (OS) based detector, referred to as the OSOMP, is proposed by using the orthogonal matching pursuit (OMP) algorithm to estimate the target support, and then projecting the compressive observations into the estimated subspace. Since OMP applies an iterative ranking process to select the components/atoms of the dictionary, the OSOMP test variable is an order statistic, which has a non-convergent distribution. To cope with this problem, a modified test statistic for the OSOMP detector is presented and an analytical expression for the probability of false alarm is obtained. We further discuss the minimum number of iterations required by the OSOMP algorithm to achieve the desired probability of detection and false alarm. Numerical simulations are conducted to verify the theoretical analysis and illustrate the performance of the proposed detector relative to several benchmark detectors. [ABSTRACT FROM AUTHOR]

• Uses the characteristic that RCS of wideband targets has various frequency response under different radiation frequencies. • The target RCS frequency response function is predicted by HMM. • In the process of joint cognitive frequency selection and power allocation, active radar transmits a set of broadband multi-frequency pulses, and uses passive radar to receive echo. It includes the information interaction between passive radar and active radar. • Joint cognitive frequency transmission and power allocation of active radar based on target RCS information. In the Low Probability of Intercept (LPI) radar system, in order to optimize the radar target detection performance, the energy consumption should be reduced while improving the target detection performance. Based on the Hidden Markov Model (HMM), this paper proposes an optimization algorithm for target detection performance joint cognitive frequency transmission and power allocation combined with the target Radar Cross Section (RCS) frequency response function prediction. The algorithm uses the target RCS's characteristics with different frequency response functions at different frequencies. The radar transmits a set of broadband pulse signals with different frequencies at the current time to irradiate the target and obtain the echo. The passive receiving system processes the echo to obtain the target RCS frequency response measurement value, and determines it as a priori information. The RCS frequency response function of the target at the subsequent time is predicted by the prior information and HMM, and the complete RCS prediction information of the target is obtained. Finally, using the radar equation, the radar transmitting power is calculated through the maximum frequency response function in the predicted values at different times and its corresponding transmitting frequency. The optimized information is then transmitted to the detection radar to complete the joint cognitive frequency transmission and power allocation. The simulation results show that the method used in this paper can lead to more accurate prediction information than the traditional algorithm, efficiently reduce the target detection transmitting power, and improve the radar target detection and LPI performance, which has a high engineering application value. [ABSTRACT FROM AUTHOR]

Passive multistatic radar is of great interest in both civilian and military applications due to its numerous advantages. A linear fusion framework for target detection in the passive multistatic radar is proposed here to improve the detection performance. In our framework, the global decision is the weighted combination of the local test statistics from spatially separated receivers, and the individual test statistics are computed through cross-correlation. A modified deflection coefficient is introduced as an optimization criterion for the selection of the weighting coefficients. Particularly, we give the weighting coefficient in closed form which is computationally efficient to calculate. In addition, we derive a closed-form expression for the probability of false alarm of this system, which is verified using Monte Carlo simulations. Numerical examples illustrate that our proposed approach outperforms its counterparts. [ABSTRACT FROM AUTHOR]

For passive radar target detection, the cross-correlation (CC) based detector is a popular method, which cross-correlates the signal received in a reference channel (RC) and the signal in a surveillance channel (SC). The CC is simple to implement and resembles the clairvoyant matched filter (MF) in idealistic conditions. However, there is limited understanding on its performance in passive sensing environments with non-negligible noise in the RC and direct-path interference in the SC. This paper examines such effects on the detection performance of the CC detector. Closed-form expressions for the probabilities of false alarm and detection of the CC detector are derived, which are employed to quantify to what extent the noise in the RC and the direct-path interference in the SC should be suppressed in order to achieve a targeted performance loss of the CC detector relative to the MF. These results are useful in designing practical CC solutions for passive radar sensing. [ABSTRACT FROM AUTHOR]

Abstract: We propose efficient target localization methods for a passive radar system using time-of-arrival (TOA) information of the signals received from multiple illuminators, where the position of the receiver is subject to random errors. Since the maximum likelihood (ML) formulation of this target localization problem is a non-convex optimization problem, semi-definite relaxation (SDR)-based optimization methods in general do not provide satisfactory performance. As a result, approximated ML optimization problems are proposed and solved with SDR plus bisection methods. For the case without position error, it is shown that the relaxation guarantees a rank-one solution. The optimization problem for the case with position error involves only a relaxation of a scalar quadratic term. Simulation results show that the proposed algorithms outperform existing methods and provide root mean-square error performance very close to the Cramer–Rao lower bound. [Copyright &y& Elsevier]

• Estimated support of the L0LS is related to the equivalent SNR of the output estimate. • The regularization parameter of the L0LS is revealed to have a clear physical meaning. • L0LS has a lower overall deviation compared with matched filtering in mild conditions. Matched filtering (MF) is widely used in radars as it has the potential to maximize the output signal-to-noise ratio (SNR). Nevertheless, when there are multiple echoes, there exists mutual interferences among these echoes in the output of the MF, which may degrade the performance dramatically, especially in the case of relatively poor waveform. To tackle this problem, we consider ℓ 0 -regularized least squares (L0LS) as an alternative to the MF in this paper. We first discuss some basic properties of the L0LS, including the relationship between the sparsity and estimation error, and the relationship between the estimated index support set and the equivalent SNR of the output estimate, etc. Then we compare the estimation deviation of the L0LS with that of the MF on the premise that the real support set is estimated by the L0LS under a certain regularization parameter. It reveals that the L0LS has a lower overall deviation compared with the MF. Finally, an example in passive radar is taken to demonstrate the potential performance improvement of the L0LS over the MF. [ABSTRACT FROM AUTHOR]

Abstract: In the context of target tracking, the Posterior Cramér–Rao Lower Bound (PCRLB) provides a powerful tool to assess the optimal achievable accuracy of target state estimation. In the bistatic configuration both geometry factors and transmitted waveform play an important role in the estimation accuracy. In this paper, we derive the PCRLB on sequential target state estimation accuracy in a bistatic radar tracking scenario, in the most general case of uncertain target measurements, i.e. when the probability of detection is less than one and the probability of false alarm is greater than zero. Then, we propose a PCRLB-based algorithm for selecting the best transmitter of opportunity for the tracking of a radar target in a multisensor Passive Coherent Location (PCL) system. [Copyright &y& Elsevier]

• Active antagonism requires the radar to take measures in advance to avoid being jammed, which can address the challenge of mainlobe jamming. • A subpulse-level frequency agile radar and a jammer that works in a transmit/receive time-sharing mode are considered. • Under the established model, Markov Decision Process is used to describe the relationship of them. • The anti-jamming strategies can be designed based on advanced deep reinforcement learning. Among different jamming techniques, mainlobe jamming is difficult to deal with for the radar and traditional passive anti-jamming methods are less effective because the angular separation between the jammer and the target is almost the same. In contrast to these passive methods, active antagonism requires the radar to take measures in advance to avoid being jammed and this can be achieved via frequency agile (FA) radar. In order to enable the FA radar to combat the jammer and obtain good performance, a deep reinforcement learning (RL) based anti-jamming strategy design method is proposed in which a transmit/receive time-sharing jammer may adopt multiple different jamming strategies. To combat the individual jamming strategy, we propose a specialized strategy learning algorithm that treats probability of detection as the reward signal and uses proximal policy optimization to solve the RL problem of the radar and the jammer. Based on the learned specialized strategies, policy distillation technique is applied to design a unified strategy which enables the FA radar to combat multiple jamming strategies. Simulation results show that the FA radar can avoid being jammed and obtain a high probability of detection whether the jammer adopts individual or multiple jamming strategies through the proposed method. [ABSTRACT FROM AUTHOR]

• A two-step generalized likelihood ratio test (GLRT) is proposed for adaptive detection in passive radars using the mixed-signal model. • The novelty of this paper is to use the structural information about the covariance and cross-covariance matrices of the decimated mixed signals for adaptive detection. • A two-step GLRT-based detector is derived in which Principal Components-Covariance Matrix Tapering (PC-CMT) technique is used to estimate the total covariance matrix of the data corresponding to multiple range cells. • The results show that that the detection performance of the two-step GLRT detector is improved compared to that of Kelly's GLRT. A two-step generalized likelihood ratio test (GLRT) is proposed for adaptive detection in passive radars using the mixed-signal model. The detector is derived based on a hypothesis test in which the spill-over of target energy into several mixed signals, and the structured correlation between the mixed signals corresponding to the adjacent range cells are considered. The detection performance is evaluated by simulations. Moreover, two upper bounds on the performance are extracted analytically. The results show that the detection performance of the detector is improved with respect to the previously proposed adaptive detection algorithms for passive radars. [ABSTRACT FROM AUTHOR]

• Signal structure is explored to improve detection performance of passive MIMO radar. • Two generalized likelihood ratio tests are derived with unknown covariance matrix. • The proposed tests are shown to be constant false alarm rate. • The proposed tests can approach the optimum performance with known covariance matrix. In passive multiple input multiple output (MIMO) radar, the transmit signals of the noncooperative illuminators of opportunity are usually not completely known. However, they are often standard communication signals with some specific signal structure. Exploiting such information, the detection performance of passive MIMO radar can be improved. In this paper we derive a generalized likelihood ratio test (GLRT) for passive MIMO radar detection when the covariance matrix of the colored Gaussian noise is unknown and the structure of the transmit signal is known but it contains some unknown information bits. Moreover, a model which employs a reasonable approximation for some practical scenarios while requiring only a limited number of training samples is also considered, and a GLRT for this model is also derived. Our tests are shown to be constant false alarm rate (CFAR), and their performance approaches the optimum performance with known signal structure and known covariance matrix when the number of training samples is increased. Finally, several numerical examples are presented to demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]