Search

Electromagnetic vector sensor (EVS) array has drawn extensive attention in the past decades, since it offers two-dimensional direction-of-arrival (2D-DOA) estimation and additional polarization information of the incoming source. Most of the existing works concerning EVS array are focused on parameter estimation with special array architecture, e.g., uniform manifold and sparse arrays. In this paper, we consider a more general scenario that EVS array is distributed in an arbitrary geometry, and a novel estimator is proposed. Firstly, the covariance tensor model is established, which can make full use of the multidimensional structure of the array measurement. Then, the higher-order singular value decomposition (HOSVD) is adopted to obtain a more accurate signal subspace. Thereafter, a novel rotation invariant relation is exploited to construct a normalized Poynting vector, and the vector cross-product technique is utilized to estimate the 2D-DOA. Based on the previous obtained 2D-DOA, the polarization parameter can be easily achieved via the least squares method. The proposed method is suitable for EVS array with arbitrary geometry, and it is insensitive to the spatially colored noise. Therefore, it is more flexible than the state-of-the-art algorithms. Finally, numerical simulations are carried out to verify the effectiveness of the proposed estimator.

Ideal array responses are often desirable to a multiple-input multiple-output (MIMO) system. Unfortunately, it may not be guaranteed in practice as the mutual coupling (MC) effects always exist. Current works concerning MC in the MIMO system only account for the uniform array geometry scenario. In this paper, we generalize the issue of angle estimation and MC self-calibration in a bistatic MIMO system in the case of arbitrary sensor geometry. The MC effects corresponding to the transmit array and the receive array are modeled by two MC matrices with several distinct entities. Angle estimation is then recast to a linear constrained quadratic problem. Inspired by the MC transformation property, a multiple signal classification- (MUSIC-) like strategy is proposed, which can estimate the direction-of-departure (DOD) and direction-of-arrival (DOA) via two individual spectrum searches. Thereafter, the MC coefficients are obtained by exploiting the orthogonality between the signal subspace and the noise subspace. The proposed method is suitable for arbitrary sensor geometry. Detailed analyses with respect to computational complexity, identifiability, and Cramer-Rao bounds (CRBs) are provided. Simulation results validate the effectiveness of the proposed method.

Vehicle target detection technology refers to the process of vehicle detection and recognition in the image data set of different backgrounds by means of feature extraction. The vehicle target detection technology based on deep learning shows obvious advantages in the accuracy and speed of target detection. With the development of science and technology, the detection and recognition of vehicles in UAV aerial images has become an important applied research direction. This paper studies the detection and recognition of UAV aerial vehicle based on deep learning, and proposes a new deep learning-based algorithm to solve the problem that incomplete vehicle targets in the UAV aerial vehicle based on YOLOV3 algorithm cannot be recognized, and vehicles close to the UAV aerial vehicle are missed. Experimental verification results show that, compared with the existing algorithms, the proposed algorithm can significantly improve the detection accuracy of UAV aerial vehicle based on deep learning while ensuring real-time performance.

Zigbee is a popular communication protocol in the Internet of things (IoT) which shows great potential in smart home. However, the smart device has the risk of being hijacked by unauthorized users and may result in privacy disclosure. Traditional device identification is based on cryptography which is easy to be cracked. Recently, radio frequency fingerprinting identification (RFFID) is popular in device identification. Traditional RFFID’s power consumption and cost is unacceptable to Zigbee. In order to reduce the cost, more effective model can be used to reduce the number of neurons. This paper proposes a RFFID method based on lightweight convolution neural network (CNN) which can adopt low power consumption and cost. The simulation result shows that this method can identification Zigbee device, and the accuracy reached 100%. Also, the parameter has reduced to about 93%.

Deterministic echo state network (ESN) models integrated with particle swarm optimization (PSO) are proposed to improve the accuracy and efficiency of stock price forecasting. ESNs have been successfully applied to financial time series forecasting because of their efficient and powerful computational ability in approximating nonlinear dynamical systems. However, reservoir construction in standard ESNs is primarily driven by a series of randomized model-building stages, because of which both researchers and practitioners have to rely on a series of trials and errors to determine parameters. An ESN with a deterministically constructed reservoir is comparable in performance to a standard ESN and has minimal complexity as well as potential for optimizations with regard to ESN parameters. In this paper, forecasting performances of the proposed PSO-DESN models are compared with those of standard ESNs for stock price prediction on the benchmark dataset of S&P 500. The comparison results demonstrate that the proposed PSO-DESNs exhibit better performance in stock price forecasting in terms of both accuracy and efficiency, thereby verifying the potential of PSO-DESNs for financial predictions.

How to identify influential nodes in complex networks is a fundamental and important problem. In this paper, we propose a novel centrality measure, called Absorption Centrality, to measure the influence of all the entities of complex networks. Based on random walks, we start with the perspective of dynamical process, instead of static structure properties which are exhibited by other traditional centrality measures, such as Degree Centrality, Betweenness Centrality and Closeness Centrality. Experimental results demonstrate the proposed Absorption Centrality is more effective than the above three centralities.

When trying to find the convex hull (CH) of a point set, humans can neglect most non-vertex points by an initial estimation of the boundary of the point set easily. The proposed CH algorithm imitates this characteristic of visual attention, starts by constructing an initial convex polygon (ICP), and measures the width and length of ICP through a shape estimation step. It then maps the point set into the new one by an affine transformation and makes most of the new points exist in a new initial convex polygon (NICP) which approximated to a regular convex polygon. Next, it discards the interior points in NICP by an inscribed circle and processes the remaining points outside NICP by Quickhull. Finally, the algorithm outputs the vertex set of CH. Two theorems are also proposed to solve an unconstrained optimization problem instead of the iteration method. Compared with four popular CH algorithms, the proposed algorithm can generate CH much faster than them and achieve a better performance.

Robust channel estimation is required for coherent demodulation in multipath fading wireless communication systems which are often deteriorated by non-Gaussian noises. Our research is motivated by the fact that classical sparse least mean square error (LMS) algorithms are very sensitive to impulsive noise while standard SLMS algorithm does not take into account the inherent sparsity information of wireless channels. This paper proposes a sign function based sparse adaptive filtering algorithm for developing robust channel estimation techniques. Specifically, sign function based least mean square error (SLMS) algorithms to remove the non-Gaussian noise that is described by a symmetric α-stable noise model. By exploiting channel sparsity, sparse SLMS algorithms are proposed by introducing several effective sparse-promoting functions into the standard SLMS algorithm. The convergence analysis of the proposed sparse SLMS algorithms indicates that they outperform the standard SLMS algorithm for robust sparse channel estimation, which can be also verified by simulation results.

According to the modern bridge health assessment and testing methods, compared with the traditional manual test to determine fracture, detection method based on digital image which applies digital image processing and pattern recognition technology to test and assess the bridge surface defect images has the characteristics of non-contact and high precision. Taking the grey-scale feature of the crack, this paper will put forward a kind of crack detection method based on multi-scale and multi-perspective. In view of the interference factors such as the holes, dirties and the others on the concrete pavement, we should analyze the characteristics of collected road surface images, and adopt DWT and NSST to resolve the source images from multiscale, and then establish grey value similarity function, and withdraw the suspected crack information according to the similarity of contrast; second, remove the false crack use connected component measurement, and transfer the problem into graph theory problem; last, use the approximation coefficient of NSST domain, the fusion approximate coefficient of DWT domain and fusion detail coefficients to inverse, transform and rebuild fusion images so as to extract real crack. Through a lot of tests on the concrete pavement picture, experimental results show that this method can achieve real pavement cracks feature extraction, and enjoys a strong practicability.

Concurrent chemoradiotherapy is the standard treatment for patients with locally advanced lung cancer. The most common dose-limiting adverse effect of thoracic radiotherapy (RT) is radiation pneumonia (RP). A randomized comparison study was designed to investigate targeted percutaneous microwave ablation at pulmonary lesion combined with mediastinal RT with or without chemotherapy (ablation group) in comparison with RT (target volume includes pulmonary tumor and mediastinal node) with or without chemotherapy (RT group) for the treatment of locally advanced non-small cell lung cancers (NSCLCs). From 2009 to 2012, patients with stage IIIA or IIIB NSCLCs who refused to undergo surgery or were not suitable for surgery were enrolled. Patients were randomly assigned to the RT group (n = 47) or ablation group (n = 51). Primary outcomes were the incidence of RP and curative effectiveness (complete response, partial response, and stable disease); secondary outcome was the 2-year overall survival (OS). Fifteen patients (31.9%) in the RT and two (3.9%) in the ablation group experienced RP (P < 0.001). The ratio of effective cases was 85.1 versus 80.4% for mediastinal lymph node (P = 0.843) and 83.0 versus 100% for pulmonary tumors (P = 0.503), respectively, for the RT and ablation groups. Kaplan-Meier analysis demonstrated 2-year OS rate of NSCLC patients in ablation group was higher than RT group, but no statistical difference (log-rank test, P = 0.297). Percutaneous microwave ablation followed by RT for inoperable stage III NSCLCs may result in a lower rate of RP and better local control than radical RT treatments.

Echo state networks (ESNs), as efficient and powerful computational models for approximating nonlinear dynamical systems, have been successfully applied in financial time series forecasting. Reservoir constructions in standard ESNs rely on trials and errors in real applications due to a series of randomized model building stages. A novel form of ESN with deterministically constructed reservoir is competitive with standard ESN by minimal complexity and possibility of optimizations for ESN specifications. In this paper, forecasting performances of deterministic ESNs are investigated in stock price prediction applications. The experiment results on two benchmark datasets (Shanghai Composite Index and S&P500) demonstrate that deterministic ESNs outperform standard ESN in both accuracy and efficiency, which indicate the prospect of deterministic ESNs for financial prediction.

Identifying influential nodes is a basic measure of characterizing the structure and dynamics in complex networks. In this paper, we use network global efficiency by removing edges to propose a new centrality measure for identifying influential nodes in complex networks. Differing from the traditional network global efficiency, the proposed measure is determined by removing edges from networks, not removing nodes. Instead of static structure properties which are exhibited by other traditional centrality measures, such as degree centrality (DC), betweenness centrality (BC) and closeness centrality (CC), we focus on the perspective of dynamical process and global structure in complex networks. Susceptible-infected (SI) model is utilized to evaluate the performance of the proposed method. Experimental results show that the proposed measure is more effective than the other three centrality measures.