In an ever-increasingly complicated electromagnetic environment with explosive radar signals density, accurate and fast recognition of dual-component has become urgent problem in the current reconnaissance system. This letter proposes a novel multi-class learning framework based on deep convolutional neural network (DCNN) for recognizing eight types randomly overlapping signals. The mainly includes preprocessing, DCNN model that aimed to extract more effective features, classification....

In this paper, we propose two kinds of improved nested arrays for direction arrival (DOA) estimation, which are obtained by rearranging physical sensors the prototype array. Specifically, sum-difference coarrays (SDCAs) proposed constructed jointly using temporal and spatial information received signal. By systematically analyzing structures arrays, derive closed-form expressions about their sensor positions, SDCAs, consecutive degrees freedom (DOFs), maximum array apertures, respectively....

Due to the polarization diversity (PD) of element patterns caused by varying curvature conformal carrier, conventional direction-of-arrival (DOA) estimation algorithms could not be applied array. In order describe PD array, parameter is considered in snapshot data model. The paramount difficulty for DOA coupling between angle information and parameter. Based on characteristic cylindrical decoupling can realized with a specially designed array structure. 2D-DOA accomplished via parallel...

Pulse repetition interval (PRI) modulation recognition and pulse sequence search are significant for effective electronic support measures. In modern electromagnetic environments, different types of inter-pulse slide radars highly confusing. There few available training samples in practical situations, which leads to a low accuracy poor effect the sequence. this paper, an approach based on bi-directional long short-term memory (BiLSTM) networks temporal correlation algorithm PRI under small...

The pattern of each element in conformal array has a difierent direction for the curvature carrier, which results polarization diversity antenna. Polarization parameters incident signals are considered snapshot data model order to describe It is required that and arrival (DOA) estimated together. An integrated frequency DOA estimation method proposed this paper cylindrical signal source obtained by constructing state-space matrix. Through well-designed conflguration elements on carriers...

Co-prime array configurations are considered attractive due to the extension of degrees freedom (DOFs) and sparse placement elements.In this paper, a 2-D direction-of-arrival (DOA) polarization estimation algorithm proposed with three-parallel co-prime sensitive which consists co-centered orthogonal loop dipole.A novel cross-covariance matrix, that not contains parameters, is constructed decouple joint problem DOA angles parameters.Then, by using vectorization operation linear...

As is well known, the prototype coprime arrays consist of two collinear subarrays, whose physical sensor number to each other. Moreover, their larger inter-element separations and array apertures make them have less mutual coupling than uniform linear arrays. Even so, there still existing in them, which has an adverse effect on direction arrival estimation. To reduce effects coupling, several improved configurations based concept difference sum coarray are proposed this paper, can be...

In this letter, a novel sparse array consisting of coprime and nested subarrays is designed corresponding two-dimensional (2-D) direction arrival (DOA) estimation method presented. The proposed can be decomposed into two planar subarrays, each which consists some identical linear arrays (NLAs). Especially, the dense in these NLAs form prototype arrays. Then, by vectorizing covariance matrices, virtual are available, have much larger apertures than physical ones. Subsequently, 2-D spatial...

Principal component analysis (PCA)-based approach for user heading estimation using a smartphone in the pocket suffers from an inaccurate of device attitude, which plays central role both obtaining acceleration signals horizontal plane and ultimate global walking direction extraction. To solve this problem, we propose novel based on two unscented Kalman filters (UKFs) fusing inertial sensors landmarks. The first UKF is developed recalibration attitude estimation. We mathematically derive...

In array signal processing systems, the direction of arrival (DOA) and polarization signals based on uniform linear or rectangular sensor arrays are generally obtained by rotational invariance techniques (ESPRIT). However, since ESPRIT algorithm relies invariant structure received data, it cannot be applied to electromagnetic vector (EVSAs) featuring circular patterns. To overcome this limitation, a fourth-order cumulant-based is proposed in paper, for joint estimation DOA EVSA. The utilizes...

Each element in the conformal array has a different pattern, which leads to performance deterioration of conventional high resolution direction‐of‐arrival (DOA) algorithms. In this paper, joint frequency and two‐dimension DOA (2D‐DOA) estimation algorithm for are proposed. The delay correlation function is used suppress noise. Both spatial time sampling utilized construct spatial‐time matrix. 2D‐DOA accomplished based on parallel factor (PARAFAC) analysis without spectral peak searching...

Nested arrays are considered attractive due to their hole-free performance, and have the ability resolve O ( N 2 ) sources with physical sensors. Inspired by nested arrays, two kinds of three-parallel subarrays (TPNAs), which composed three parallel sparse linear different inter-element spacings, proposed for two-dimensional (2-D) direction-of-arrival (DOA) estimation in this paper. We construct cross-correlation matrices combine them as one augmented matrix first step. Then, vectorizing...

Accuracy performance of WiFi fingerprinting positioning systems deteriorates severely when signal attenuations caused by human body are not considered. Previous studies have proposed based on user orientation using compasses built in smartphones. However, always cannot provide required accuracy estimation due to the severe indoor magnetic perturbations. More importantly, we discover that only orientations but also smartphone carrying positions may affect greatly. Therefore, propose a novel...

The electronic support measure has a pivotal role in warfare. Radar signal deinterleaving achieves complex radar pulse streams to deinterleave by analysing the intercepted emitter signals and estimating repetition interval (PRI) values. At present, most of algorithms use classical model pre‐sorting combined with PRI algorithm, which is challenging adapt current environment. agility parameters increases error rate; existence jitter missing leads sub‐harmonic problem low estimation accuracy;...

Abstract In bearings-only localization, clustering-based methods have been widely used to remove spurious intersections by fusing multiple bearing measurements from different observation stations. Existing clustering methods, including fuzzy C-mean (FCM) and density-based spatial of applications with noise (DBSCAN), must specify the number clusters threshold for defining neighborhood density, respectively, which are always unknown difficult estimate. Moreover, in dense radiation source...

Heading estimation using inertial sensors built‐in smartphones has been considered as a central problem for indoor pedestrian navigation. For practical daily lives, it is necessary heading to allow an unconstrained use of smartphones, which means the varying device carrying positions and orientations. As result, three special human body motion states, namely, random hand movements, position transitions, user turns, are introduced. However, most existing approaches neglect may render large...

The cardinalized probability hypothesis density (CPHD) filter is an alternative approximation to the full multi-target Bayesian for tracking multiple targets. However, although joint propagation of posterior intensity and cardinality distribution in its recursion allows more reliable estimates target number than PHD filter, CPHD suffers from spooky effect where there exists arbitrary mass shifting presence missed detections. To address this issue Gaussian mixture (GM) implementation paper...

Abstract Automatic radar modulation recognition plays a significant role in both civilian and military applications. With the rapid development of deep learning, convolutional neural networks have achieved demonstrated success signal recognition. However, usually only recognise trained classes, when dataset changes, need to be retrained. actual applications, model needs predict new signals, size training set will continue accumulate. Therefore, few‐shot learning on dynamic datasets become...

In an ever-increasingly complex electromagnetic environment, automatic radar signal recognition is becoming vital. Convolutional neural networks have been widely used for recognition, but deep learning-based algorithms only recognize trained classes. Recognizing novel signals with few-shot samples in open environment still a challenging research problem. this letter, learning algorithm based on the tensor imprint and convolutional classification layer proposed can avoid spatial information...

This paper studies the dynamic estimation problem for multitarget tracking. A novel gating strategy that is based on measurement likelihood of target state space proposed to improve overall effectiveness probability hypothesis density (PHD) filter. Firstly, a measurement-driven mechanism this technique designed classify measurements. In mechanism, only measurements existing targets are considered in update step while newborn used exploring targets. Secondly, enables development heuristic...

In Bayesian multi-target tracking (MTT), knowledge of clutter intensity is required for effective state estimation. this paper, we propose an online filter that can operate under background with unknown intensity. Our solution based on the Poisson multi-Bernoulli mixture (PMBM) jointly estimating and rate. The rate modeled as Gamma distribution, consequently, derived PMBM recursion adapts remains closed. Moreover, adopt a Gibbs sampler to find finite number global hypotheses, then...