Search

The present study provides a wideband high-gain monopulse antenna based on a dielectric lens operating in X- and Ku-bands, in which a wideband dielectric lens is designed and employed to fulfill the radiation pattern and bandwidth necessities of a monopulse antenna. The proposed configuration has four horns allowing for the simultaneous creation of $\Delta $ and $\Sigma $ designs in two perpendicular planes. The main advantages of the proposed dielectric lens are low cost, lightweight, and easy fabrication using 3-D printing technology. The measurement findings show that the peak gain of the sum pattern is 28.9 dBi with a peak aperture efficiency of 60% over the desired frequency bandwidth. The suggested design can produce a simultaneous sum and two distinct difference patterns in orthogonal planes, meeting the rigorous demands for speed and accuracy in tracking applications.

Sulfur hexafluoride (SF6) is the most widely used gas insulation dielectric in power system, Currently, SF6 degradation technology based on non-thermal plasma faces issues with high content of difficult-to-treat byproducts like SO2F2. This study proposes improving the SF6 destruction and removal efficiency (DRE) through tandem DBD degradation and investigates the impact of adding H2 on the tandem DBD degradation of SF6 and its decomposition products. The results show that under conditions of 2% initial SF6 concentration, 1.5% H2, 150 mL/min gas flow rate, and 80W input power, the DRE in the first-stage tube is 73.2%, reaching 78.02% in the second-stage tube, and the DRE of SF6 in the dual-stage tandem reaches as high as 82.82%. In terms of decomposition products, there is no difference in the types of products formed in the two reaction tubes when H2 is added, and the main products are SO2F2, SO2, SOF2, and SOF4. The addition of H2 will inhibit the formation of S-O-F gas products, improves DRE and significantly regulates the distribution of degradation products, reducing the amount of SO2F2 generated in the second-stage tube and making SO2 the main product. Adding H2 with a concentration of 2.5 % can avoid the formation of SO2F2 in the second-stage tube. This study provides guidance for efficient SF6 degradation and product control.

This paper investigates a reconfigurable intelligent surface (RIS)-assisted unmanned aerial vehicle (UAV)-enabled wireless powered communication network (WPCN). In the system, a UAV acts as a hybrid access point (HAP) to charge users in the downlink (DL) and receive messages in the uplink (UL). In particular, the RIS is exploited to significantly enhance the efficiency of both the DL and UL transmission. Our objective is to enhance the minimum throughput among all ground users by jointly optimizing the horizontal location of UAVs, the transmit power of users, transmission time allocation, and passive beamforming vectors at the RIS. To address this problem, we present an alternating optimization-based algorithm with low complexity to decompose the problem into four subproblems and solve them sequentially. In particular, we derive a lower bound of the composite channel gain to tighten the constraints and employ successive convex approximation (SCA) to optimize the horizontal location of the UAV. The transmit power closed-form optimum solutions are then obtained, and the problem of time allocation is reformulated as a linear programming problem. Finally, we optimize the passive beamforming vectors by adopting semi-definite relaxation (SDR). The effectiveness of the algorithm is supported by numerical results, which also demonstrate that the RIS-assisted UAV-enabled WPCN outperforms the traditional WPCN in terms of the minimum throughput.

In light of the inadequacies of traditional P-wave arrival picking algorithms using long and short windows, which exhibit poor anti-noise ability and do not meet the requirements for efficient and accurate P-wave arrival picking, this study proposes a CGAS deep learning algorithm for P-wave arrival picking in mine micro-seismic events. The algorithm constructs the depth characteristics of microseismic waveform through deep learning training and converts the problem of picking up the microseismic waveform into a classification problem by employing data set segmentation and classification. This allows the algorithm to pick up the characteristic information of microseismic waveform and improve the accuracy of P-wave picking. The algorithm was applied to pick up the three-component micro-seismic waveforms obtained from the microseismic monitoring system in a mining area in Liaoning from 2019 to 2020. Through ablation experiments, it was found that adding each structure to the pickup model improved the relevant indexes to some extent. When all network structures were used, the accuracy of the P-wave pick-up model reached 98.61%, ensuring the accuracy of P-wave pickup upon arrival and demonstrating the effectiveness of each layer structure in the classification of P-wave arrival. Compared to other P-wave arrival picking algorithms such as STA/LTA, U-Net++, Dpick, PhaseNet and EQTransformer, the algorithm proposed in this paper exhibited high precision in P-wave arrival time picking, providing a new research idea and technical means for efficiently determining the P-wave arrival of events.

As an interdisciplinary course, Machine Vision combines AI and digital image processing methods. This paper develops a comprehensive experiment on forest wildfire detection that organically integrates digital image processing, machine learning and deep learning technologies. Although the research on wildfire detection has made great progress, many experiments are not suitable for students to operate. Also, the detection with high accuracy is still a big challenge. In this paper, we divide the task of forest wildfire detection into two modules, which are wildfire image classification and wildfire region detection. We propose a novel wildfire image classification algorithm based on Reduce-VGGnet, and a wildfire region detection algorithm based on the optimized CNN with the combination of spatial and temporal features. The experimental results show that the proposed Reduce-VGGNet model can reach 91.20% in accuracy, and the optimized CNN model with the combination of spatial and temporal features can reach 97.35% in accuracy. Our framework is a novel way to combine research and teaching. It can achieve good detection performance and can be used as a comprehensive experiment for Machine Vision course, which can provide the support for talent cultivation in machine vision area.

Although increased attention is being given to understanding how people search to build task outcomes, a formal model of the relation between how people search and how people build task outcomes is still lacking. This paper proposes a unified probabilistic model of how people search to build outcomes. The model involves 3 types of searcher behaviors (i.e., query submission, document selection, and information transformation) to model the effect of the information collected during search, and uses the item response theory to capture the ternary relations between the ability to transform information, the information collected, and the probability of successfully building task outcomes. We evaluate the proposed model in the task of identifying searchers’ proficiencies under the assumption that high proficiency searchers would have high abilities to transform information. The results obtained high accuracies and F1 scores, which could reflect the effectiveness of the proposed model. The model contributes to the formal understanding of how people search to build task outcomes, and provides new possibilities for personalized and session-based information retrieval research.

Power-line communication (PLC) is a promising communication technology of constructing IoT (internet of things) systems, and it employs IEEE 1901 as its medium access control (MAC) protocol. Current studies about IEEE 1901 MAC performance analysis do not consider the influence of physical layer's retransmission protocols, i.e., ARQ (automatic repeat request) and HARQ-CC (hybrid ARQ with chase combining). Moreover, how 1901 protocol affect the energy consumption in PLC networks, and whether 1901 protocol has bound performance are also unsolved in these existing works. Focusing on the above problems, we put forward a cross-layered theoretical model to insightfully analyze 1901 protocol, where the impacts of physical layer's retransmission protocols (ARQ and HARQ-CC), channel fading modes and practical configurations (buffer size, finite traffic load, etc.) are comprehensively taken into consideration in the modeling process. On this basis, we derive the closed-form expressions of 1901 MAC metrics considering retransmission protocols. Furthermore, we construct an energy consumption model, and provide the bound performance analysis for 1901 protocol. Finally, we evaluate the performance of 1901 protocol under different retransmission protocols and system parameters, and verify the proposed cross-layered theoretical model.

In this paper, a novel soft computing algorithm is designed for the numerical solution of third-order nonlinear multi-singular Emden–Fowler equation (TONMS-EFE) using the strength of universal approximation capabilities of Legendre polynomials based Legendre neural networks supported with optimization power of the Whale Optimization Algorithm (WOA) and Nelder-Mead (NM) algorithm. Unsupervised error functions are constructed in terms of mean square error for governing TONMS-EF equations of first and second order. Unknown designed parameters in LeNN structure are optimized initially by WOA for global search while NM algorithm further enhances the rapid local search convergence. The proposed algorithm’s objective is to show the accuracy and robustness in solving challenging problems like TONMS-EFE. To study our designed scheme’s performance and effectiveness, LeNN-WOA-NM is implemented on four cases of TONMS-EFE. The results obtained by the proposed algorithm are compared with the Particle Swarm Optimization (PSO) algorithm, Cuckoo search algorithm (CSA), and WOA. Extensive graphical and statistical analysis for fitness value, absolute errors, and performance indicators in terms of mean, median, and standard deviations show the proposed algorithm’s efficiency and accuracy.

High-resolution (HR) Mars images have great significance for studying the land-form features of Mars and analyzing the climate on Mars. Nowadays, the mainstream image super-resolution methods are based on deep learning or CNNs, which are better than traditional methods. However, these deep learning based methods obtain low-resolution(LR) images usually by using an ideal down-sampling method (e.g. bicubic interpolation). There are two limitations in the existing SR methods: 1) The paired LR-HR data by using such methods can achieve a satisfactory results when tested on an ideal datasets. But, these methods always fail in real Mars image super-resolution, since real Mars images rarely obey an ideal down-sampling rule. 2) The LR images obtained by ideal down-sampling methods have no noise while real Mars images usually have noise, which leads to the super-resolved images are not realistic in texture details. To solve the above-mentioned problems, in this article, we propose a novel two-step framework for Mars image super-resolution. Specifically, to address limitation 1), we focus on designing a new degradation framework by estimating blur-kernels. To address limitation 2), a Generative Adversarial Network (GAN) is trained to generate noise distribution. Extensive experiments on the Mars32k dataset demonstrate the effectiveness of the proposed method, and we achieve better qualitative and quantitative results compared to other SOTA methods.

Continuous emission monitoring system (CEMS) is widely used in the field of environmental protection, but the maintenance routines of it consist mostly of pre-defined activities. Reliability-centered maintenance (RCM) uses a systematic approach to develop maintenance strategies that take into account factors such as system life, operating efficiency, and maintenance cost-effectiveness. This paper is motivated to apply RCM to sampling subsystem in CEMS. Based on the failure mode and effect analysis (FMEA), the RCM logic decision used to identify the reasonable maintenance mode is implemented. Subsequently, an average availability model is proposed to determine the maintenance interval cycle. Finally, the maintenance outline is formed, which can be then further used as valuable information to improve equipment reliability and develop maintenance strategies.

In this paper, a wideband diplexer with narrow channel spacing is designed by using hybrid bandpass-bandstop structures. Two wideband bandpass structures are designed with the transmission zeros at the upper or lower passband edges to achieve high skirt selectivity. Two bandstop structures based on half-wavelength coupled lines are integrated to the two bandpass structures to introduce additional transmission zeros and thus the skirt selectivity is further improved. Accordingly, two channel filters with high skirt selectivity and high stopband rejection are designed to realize a wideband microstrip diplexer with very narrow channel spacing. For verification, a wideband diplexer operating at 1.71-2.17 GHz and 2.30-2.70 GHz is implemented, which covers multiple frequency bands for different mobile systems. The measured results show excellent performance of passband flatness, high in-band isolation of better than 35 dB and low minimum insertion losses of 0.46 and 0.50 dB for the two channels.

This paper presents a broadband high-efficiency harmonic-tuned power amplifier (PA) with quasi-elliptic low-pass responses. A combination of continuous Class-F-1 and extended continuous Class-F PA modes is employed to significantly expand the design space. A quasi-elliptic low-pass matching network is proposed, which can realize a broadband impedance matching in the predefined optimal impedance region desired by the combination of PA modes. Furthermore, two transmission zeros are generated near the passband, exhibiting high skirt selectivity and providing rapid impedance transition from the passband to the stopband. A wide stopband covering up to the third harmonic is achieved which shows good harmonic suppression. Design procedures of the proposed broadband PA are presented. To verify the proposed methodology, the broadband PA is fabricated and measured. The implemented PA achieves a bandwidth of 145.2% from 0.5 to 3.15 GHz. Over this frequency range, the drain efficiency is measured as 58-74.9% with the output power of greater than 39.03 dBm and a large signal gain ranging from 8.43 to 15.67 dB. A wide stopband is realized from 3.4 to 10 GHz, showing excellent quasi-elliptic low-pass responses. The measured adjacent leakage ratios (ACLRs) using a 20-MHz LTE signal with digital pre-distortion are below -45.06 dBc.

This paper presents a circularly polarized (CP) rectenna with the advantage of achieving high efficiency in both wide operating power and frequency ranges. The proposed rectenna is composed of a high-efficiency rectifier and broadband CP antenna. In the proposed rectifier, a novel wideband resistance compression technique is presented, highly improving the matching performance of the circuit in a wide range of input power and frequency. The technique is achieved by an impedance manipulation network and a coupled-lines-based resistance compression network (RCN). Theoretical analysis is carried out and closed-form equations are derived for the design of the rectifier. Simulated and measured results show that over 60% (up to 76%) conversion efficiency is achieved for the input power range of 5-17 dBm and frequency band of 1.7-2.9 GHz (mobile, Wi-Fi and ISM bands). Subsequently, unlike the conventional half-wavelength slot antenna, a compact CP antenna with a side length of quarter wavelength is designed by using a 1-wavelength loop slot and a coupled-line-based phase shifter. It has wide axial-ratio bandwidth and good radiation pattern. At last, by integrating the proposed rectifier together with the CP slot antenna, a broadband CP rectenna is completed with higher efficiency by about 18.6% than that without wideband RCN.

The incoming 5G technology is expected to proliferate tremendous internet-of-thing (IoT) services with real-time and mobility requirements, which are quite different from the legacy cloud services. Due to the centralized management relying on distant datacenters, cloud computing is short of satisfying the stringent IoT requirements, such as ultra-low latency, mobility, etc. Instead, distributed edge computing, such as fog computing has been coined as a promising approach and has received enormous attention in recent years. In this paper, to optimally provision the huge volume of IoT services with significant diversity, we propose to efficiently organize the leisure network devices in the network edge to form fog networks (fogs), which are then integrated with the cloud to provide storage and computing resources. Specifically, we propose two Integer Linear Programming (ILP) models to solve the fog planning issue under the integrated Cloud-Fog (iCloudFog) framework. In the first ILP model, the objective is to minimize the CAPEX cost caused by planning fogs and the OPEX cost caused by utilizing the planned fogs. In the second ILP model, the objective is to minimize the power consumption while maximizing the number of successfully provisioned IoT tasks on the planned fogs. The proposed ILP models are numerically evaluated by considering different IoT task requirements, such as real-time and mobility. The numerical results show that efficiently planned fogs can help to reduce the planning overhead while satisfying diverse IoT task requirements.

Maximum a posteriori (MAP) approach, based on Bayesian criterion, is proposed to overcome the low azimuth resolution in real-aperture imaging. The essence of this approach is to use the statistical characteristics of the imaging background and target to invert the real target scene. This paper presents a deconvolution method based on Maximum a posteriori (MAP) criterion, which combines the Rayleigh distribution and Lognormal distribution, to realize high angular resolution for sea-surface target. Firstly, Rayleigh distribution is considered to express the statistical properties of sea clutter. Moreover, the Lognormal distribution is employed to represent the statistical properties of target as prior information. The reason is that Lognormal distribution can be approximatively regarded as a combined constraint term. Finally, the optimization theory is utilized to obtain the iterative estimated solution. The processed results of simulation and measured data are given to verify the performance of proposed algorithm.

With the rapid development of communication and network technologies, novel information services and applications are rapidly growing worldwide. Advanced communications and networks greatly enhance the user experience, and have a major impact on all aspects of people’s lifestyles in terms of work, society, and the economy. Although advanced techniques have extensively improved users’ quality of experience (QoE), they are not adequate to meet the various requirements of seamless wide-area coverage, high-capacity hot-spots, low-power massive-connections, low-latency and high-reliability, and other scenarios. Therefore, it is a great challenge to develop smart communications and networks that support optimized management, dynamic configuration, and feasible services.

To meet the requirement for the exact computation of rotational stiffness, this paper puts forward a mesh adaptation based analysis method for a compact multi-throttle aerostatic guideway and its constituent micro-structure. Contrary to previous methods that substituted or even ignored them, the dimensionless thickness pertaining to the depth of the micro-structure is restored and incorporated into each computational grids. The mesh adaptation procedures accept the dimensionless thickness as an intermediate variable for the target mesh and apply local modifications to adapt the mesh to match the specified size. Then, the varying regularity of rotational stiffness by micro-structures can be studied by k-ε turbulence model when tilt. The tilt angle, although small, leads to a wedge-shaped film and then an uneven pressure distribution which has been also taken into consideration. Guided by the mesh adaptation method, a vertical compact guideway with high rotational stiffness of 7.22 × 103 Nm/rad is designed and manufactured for photolithography. The experiments verify its practical accuracy and effectiveness.

With the explosive growth of image big data in the agriculture field, image segmentation algorithms are confronted with unprecedented challenges. As one of the most important images segmentation technologies, the fuzzy c-means (FCMs) algorithm has been widely used in the field of agricultural image segmentation as it provides simple computation and high-quality segmentation. However, due to its large amount of computation, the sequential FCM algorithm is too slow to finish the segmentation task within an acceptable time. This paper proposes a parallel FCM segmentation algorithm based on the distributed memory computing platform Apache Spark for agricultural image big data. The input image is first converted from the RGB color space to the lab color space and generates point cloud data. Then, point cloud data are partitioned and stored in different computing nodes, in which the membership degrees of pixel points to different cluster centers are calculated and the cluster centers are updated iteratively in a data-parallel form until the stopping condition is satisfied. Finally, point cloud data are restored after clustering for reconstructing the segmented image. On the Spark platform, the performance of the parallel FCMs algorithm is evaluated and reaches an average speedup of 12.54 on ten computing nodes. The experimental results show that the Spark-based parallel FCMs algorithm can obtain a significant increase in speedup, and the agricultural image testing set delivers a better performance improvement of 128% than the Hadoop-based approach. This paper indicates that the Spark-based parallel FCM algorithm provides faster speed of segmentation for agricultural image big data and has better scale-up and size-up rates.

A compact rectifier with high efficiency at wide input power dynamic range is proposed in this paper. The circuit consists of a matching network, two shunt connected diodes with an integrated impedance compression network (IICN) and a dc-pass filter. The IICN significantly reduces the diode impedance variation when input power changes, hence improves matching as well as efficiency of the rectifier at wide input power dynamic range. Different from the conventional resistance/impedance compression networks, which can only work under certain impedance condition with dual-branch topology, the IICN can directly operate two diodes in parallel in a single-branch with no impedance limitation. Thus, less components and branch are needed, resulting in size reduction and lower circuit complexity. The design flexibility is also enhanced. Mechanism of the IICN is analyzed. A prototype operating at 2.45 GHz is optimized, fabricated and measured. The total size of the prototype is 0.18λ2. The efficiency higher than 70% of the peak value is realized at input power ranging from 0.8 dBm to 18.0 dBm.

Metro door system (MDS) is one of the most important and fault-prone parts of metro train. Reliability-centered maintenance (RCM) is a systematic engineering process for determining the preventive maintenance requirements of equipment and optimizing the maintenance system. This paper is motivated to apply RCM to the maintenance of metro door system. Firstly, the subsystems and components of MDS are introduced. Then a fuzzy FMECA method is proposed to define the hazard degree of 10 failure modes from 5 components with high failure rate to conduct reliability analysis. Finally, reliability-centered maintenance strategy is made after maintenance mode decision and calculation of maintenance interval cycle, which provides theoretical basis to the maintenance of the metro train.

Doppler beam sharpening (DBS) technology is widely used in applications, such as helicopter rescue and early warning surveillance. To obtain the desired DBS images with high quality, accurate Doppler centroid estimation (DoCE) is necessary. Conventional methods for Doppler centroid estimation based on navigational devices are sensitive to the errors of the measured motion parameters. Hence, several alternative data-depended approaches have been developed to reduce the error. In this paper, a novel data-depended Doppler centroid estimation method is proposed to improve the image quality of DBS. We begin the method by analyzing the characteristics of range-Doppler distribution in different regions of interests. Then, the edge feature of range-Doppler distribution in forward-looking direction is extracted using morphological filtering and edge detection methods. We will show that the edge feature defines the required Doppler centroid parameters, which can be utilized to estimate the Doppler centroids of the full scene. At last, the estimation error is reduced through fitting the edge with the minimum mean square error (MMSE) algorithm. As compared with conventional Doppler centroid estimation methods, the proposed method can significantly provide reliable estimation accuracy under low echo signal to noise ratio, independent of conditions that strictly required by conventional methods. Simulations and experiments verify the proposed method.

Mining interactions between chemicals and proteins/genes is of crucial relevance for clinical medicine, adverse drug effects, and pharmacological research. Although chemical-protein interactions (CPIs) can be manually extracted, this process is expensive and time-consuming. Therefore, it is of considerable significance to automatically extract CPIs from biomedical literature. Currently, the popular methods for CPI extraction are based on deep learning to avoid sophisticated handcrafted features derived from linguistic analyses. However, the performance of existing methods is usually unsatisfactory. The reasons may be that (1) traditional word-embedding methods cannot adequately model context information, and (2) it is difficult to effectively distinguish which words play critical roles in long biomedical sentences. In this study, we propose a novel Deep-contextualized Stacked Bi-LSTM model (DS-LSTM) to tackle the drawbacks of existing methods. Specifically, our model mainly consists of three components: deep contextualized word representations, the entity attention mechanism, and stacked bidirectional long short-term memory networks (Bi-LSTMs). The deep contextualized word representations are introduced to effectively model complex characteristics of word use (e.g., syntax and semantics) and the variations of these words in the context (i.e., to model polysemy), thereby generating context information. The entity attention mechanism is applied to prioritize the weights of words associated with target entities to distinguish which words play critical roles in long biomedical sentences. We evaluate our model on the CHEMPROT corpus. Our approach achieves a micro-averaged F-score of 69.44%, which is significantly higher than existing state-of-the-art methods. Experimental results show that our approach can adequately model context information, effectively distinguish which words play critical roles in long biomedical sentences and, therefore, improve the overall performance.

This paper presents a compact dual-band ±45° dual-polarized base-station array based on embedded filtering elements. The proposed antenna array consists of 9 dual-band antenna units, which operate at the lower band (LB) of 790-862 MHz and the higher band (HB) of 880-960 MHz with a small frequency spacing. In each unit, one antenna element is embedded into the other one, featuring a very compact size. To alleviate the serious cross-band mutual coupling between the two elements, the LB patch antenna generates a boresight radiation null at its higher band edge, and the HB element is designed with a band-stop response at its lower band edge. In this way, high cross-band isolation between the two embedded elements can be obtained. For demonstration, a compact dual-band ±45° dual-polarized filtering antenna array with 9 units is designed and fabricated to fit the specification of LB and HB bands (|S11|

In recent years, with the rapid growth of Artificial Intelligence (AI) and the Internet of Things (IoT), the question answering systems for human-machine interaction based on deep learning have become a research hotspot of the IoT. Different from the structured query method in traditional Knowledge Base Question Answering (KBQA) systems based on templates or rules, representation learning is one of the most promising approaches to solving the problems of data sparsity and semantic gaps. In this paper, an answer acquisition method for KBQA systems based on a dynamic memory network is proposed, in which representation learning is employed to represent the natural language questions that are raised by users and the knowledge base subgraphs of the related entities. These representations are taken as inputs of the dynamic memory network. The correct answers are obtained by utilizing the memory and inferential capabilities. The experimental results demonstrate the effectiveness of the proposed approach.

The formula for analyzing the response characteristics of the induced current on a transmission line (TL) excited by high-frequency electromagnetic (EM) waves is a mixed integral-differential equation. The integral part is the Fredholm integral equation of the first kind with serious ill-posedness, which causes difficulty in solving it stably. The regularization technique is a useful method for the analysis of the ill-posed problems. Therefore, in this paper, a method based on the Tikhonov regularization technique is proposed to analyze the high-frequency electromagnetic response of a finite length TL illuminated by external EM waves. The integral operator is discretized by the Simpson formula. The validity and correctness of the proposed method have been verified by numerical examples and the applicability of the proposed method at different errors has been analyzed via the TL with different height.

The problem of low azimuth resolution has restricted the applicability for radar forward-looking imaging in practice. In this paper, a sparse with fast majorization-minimization (SFMM) superresolution algorithm was proposed to realize fast superresolution imaging of sparse targets in radar forward-looking area. First, we analyzed the azimuth signal of the radar forward-looking area and modeled the azimuth signal as a convolution of antenna pattern and targets distribution. Second, the superresolution problem was converted into an L1 regularization issue by introducing the L1 norm to represent the distribution of the targets under the regularization framework. Third, according to the principle of majorization-minimization (MM) algorithm, a simple L2 regularization issue was obtained to replace the difficult L1 one, and the real target distribution was obtained by solving the L2 regularization problem (We named it sparse with MM (SMM) superresolution algorithm for convenience). Then, in order to improve the computational efficiency of the algorithm, we adopted the second-order vector extrapolation idea to accelerate the conventional MM algorithm and solve the L2 regularization problem. The simulation and real data verified that the proposed SFMM algorithm not only improves the azimuth resolution in radar forward-looking imaging but also increases convergence speed on the basis of SMM superresolution algorithm.

A novel simulation technique is proposed for the radiation properties of time-varying media. This method divides the time-variation process into finite discrete time steps. The radiation field of the media at each instant in time is simulated by using commercially available frequency-domain solvers, and then these results are processed and synthesized to obtain the radiation field of the time-varying media. In other words, by using this technique, the radiation field of the time-varying media can be obtained based on the frequency-domain simulation results. The proposed technique is introduced with a time-varying grating, and the scattered wave and radiation pattern are illustrated. Moreover, an analytical model of the time-varying grating is also introduced to further explore the proposed technique. This simulation technique can be used to analyze many applications such as the radiation properties of time-modulated media, especially space-time-modulated metamaterials. It fills the gap of existing simulation methods and provides a new way for direct EM simulation of time-varying structures.

Recently, the time-updating $q$ -norm sparse covariance-based estimator ( $q$ -SPICE) was developed for online spectral estimation of stationary signals. In this work, this development is furthered to deal with non-stationary signals. By introducing a weighting matrix defined by a forgetting factor, the generalized least absolute shrinkage and selection operator (LASSO) is generalized, in order to allow for changes in the spectral content. As shown here, the resulting LASSO formulation can be solved in a simple manner using cyclic minimization, enabling recursive estimation for non-stationary signals. The proposed generalized time-updating $q$ -SPICE offers the same benefits as the original estimator, including being computationally efficient at constant computational and storage cost, but also allows for substantial improvements when dealing with non-stationary signals. The performance of the method is evaluated using both stationary and non-stationary signals, indicating the preferable performance of the generalized formulation as compared to the original time-updating SPICE algorithm.

In this paper, we study secrecy simultaneous wireless information and power transfer (SWIPT) in downlink non-orthogonal multiple access (NOMA) systems comprising a base station (BS), multiple information receivers (IRs), and multiple energy receivers (ERs) that have potential to wiretap the IRs. The goal of this paper is to maximize the sum secrecy rate (SSR) of the system subject to the individual IR's minimum data rate requirement and the individual ER's minimum harvested energy requirement. The corresponding problem involves resource allocation via max-min function, which is non-convex and difficult to solve directly. In order to tackle this, we first transform the original non-convex problem to a sequence of convex sub-problems which can be solved simultaneously. Hence, a closed-form solution of the optimal power allocation policy is derived based on the Karush-Kuhn-Tucker conditions. Numerical results validate the theoretical findings and demonstrate that the significant performance gain over the orthogonal multiple access scheme in terms of SSR can be achieved by the proposed algorithm in a SWIPT-enabled NOMA system.

Group division is a typical task involved in online social network analysis. Group division can be based on static and dynamic relationships. Compared to static group division, dynamic interactions between users can better reflect the intimacy of relationships between users in the online social network. This paper proposes a group division method based on the analysis of interactive behaviors among users and constructs a single-dimensional network structure between users based on behaviors of connecting, commenting, forwarding, and liking among users. Then, several single-dimensional networks are integrated into one complex network, this complex network is divided into groups and different groups of the online social network are identified. The experimental results show that the group division method based on users and behaviors is effective. The method proposed in this paper is used to identify Tibetan user groups and provides a theoretical basis for analyzing the health statuses of Tibetan user groups from online social networks.

Although the preclinical detection of Parkinson's disease (PD) has been explored, a practical, inexpensive, and overall screening diagnosis has yet to be made available. However, due to the large variability and complexity in progress of PD and the difficulties in gathering a single time-point measurement of a single sign, the goal of precision treatment and assessment severity would be impossible to achieve. Hence, the repeated monitoring and tracking of individuals during their daily living activities at different times would also be of great importance for treating this chronic disease. We propose a deep multi-layer perceptron (DMLP) classifier for behavior analysis to estimate the progression of PD using smartphones. This paper aims to identify severity in PD patients' actions by analyzing their speech and movement patterns, as measured with a smartphone accelerometer in their pocket at different times of the day. Popular machine learning classification algorithms, such as logistic regression, random forests, k-nearest neighbors, M5P, and DMLP, are applied on one dataset from the University of California Irvine and another dataset collected by the authors to classify each patient as being Parkinson positive or negative. We further measure the success of each method for their ability to correctly classify the patients into one of these categories. Of the experimental models, it is demonstrated that DMLP performs the best in both datasets.

In this paper, a triple-band impedance matching network is introduced and applied to rectifier design. It can be used to realize impedance matching at three frequencies simultaneously. Theoretical analysis of the triple-band matching network is carried out, and a rectifier operating at 0.85, 1.77, and 2.07 GHz is designed, fabricated, and measured. Measurement results show that when the input power is 0 dBm, the measured maximum RF-to-dc power conversion efficiencies are 61.9%, 71.5%, and 60.5% at the frequencies of 0.85, 1.77, and 2.07 GHz, respectively. Compared with other related designs, the proposed rectifier features high efficiency and compactness. Thus, it is suitable for ambient RF energy harvesting application.

Medical Imaging is a technique to create visual representations of the interior of the body, with the aim of making accurate diagnoses and optimized treatments. Many medical imaging techniques are widely used to produce images, such as computer tomography (CT), ultrasound (US), positron emission tomography (PET), single photon emission computed tomography (SPECT), and magnetic resonance imaging (MRI)/functional MRI (fMRI).

Patch antennas with multiple feeds are proposed in this paper. The antennas consist of a traditional square patch and feeding probes that are set in a line orthogonal to the polarization direction. In addition to radiating EM energy just like the conventional antennas, the proposed multi-port antennas can also realize on-antenna power combining without power combiner. Thanks to this, high-line losses and large system size caused by extra power combiner can be avoided. Moreover, since the location and the number of ports have a limited effect on the antenna performances, flexible feeding strategies can be chosen in different applications. For demonstration, four patch antennas with a different number of ports (one port, two ports, three ports, and four ports) are designed at the center frequency of 1.5 GHz. Compared with the traditional one-port patch antenna, the proposed three multi-port patch antennas can achieve nearly the same antenna performance, including bandwidth, polarization, efficiency, and radiation pattern. Besides, one-port antenna and two-port antenna integrated with a two-way power amplifier are designed and measured, respectively. Then, they are compared to further prove the proposed power combining method in multi-port antennas.

In this paper, a narrowband filtering switch with low loss and high selectivity is presented based on coaxial resonators for the first time. PIN diodes mounted on the printed circuit boards are embedded into a coaxial filter to enable ON and OFF states. In the ON-state, the PIN diodes are turned OFF, which do not introduce the loss and affect the linearity. Two transmission zeros are generated by a novel feeding structure, which improves the skirt selectivity. In the OFF-state, the PIN diodes are turned on. Then, lumped capacitors are loaded to the coaxial resonators so that the resonant frequencies of the resonators are changed. The passband at the operating frequency cannot be formed, resulting in high isolation. For demonstration, the coaxial-resonator-based filtering switch is designed and fabricated. Good agreement between simulated and measured results verifies the proposed ideas. Comparison with other reported filtering switches is given. The proposed filtering switch shows the advantages of high Q-factor, relatively compact size, and wide stopband responses, which is attractive in wireless systems.

With the explosion of big data technology, healthcare data is continuously and rapidly growing, with abundant and various values. There are wide varieties of data and heterogeneous healthcare data (images, text, video, raw sensor data, etc.) that are generated and required to be effectively stored, processed, queried, indexed and analyzed. These datasets differ widely in their volume, variety, velocity and value, including patient-oriented data such as electronic medical records (EMR), public-oriented data such as public health data, and knowledge-oriented data such as drug-to-drug, drug-to-disease, and disease-to-disease interaction registries. Big data in healthcare brings great challenges but plays an important role in healthcare transformation. The traditional techniques do not compromise end-users’ Quality of Service (QoS) in terms of data availability, data response delay, etc. It is urgent to develop software tools and techniques that support rapid query processing and speed-up data analytics, which provide awareness and knowledge in real-time.

A compact filtering switch with wide-stopband responses is presented in this paper. The resonator is designed using microstrip transmission lines and lumped capacitors. It is utilized to shift up the second harmonic frequency for wide-stopband performance and miniaturize the circuit size. In the ON-state, the circuit is the same as the bandpass filter without signals passing through the p-i-n diode, which avoids additional loss introduced by p-i-n diode. Two transmission zeros are generated near the passband to enhance the skirt selectivity. In the off-state, the signals from port 1 to port 2 passing through different transmission path are cancelled out, resulting in high isolation. For demonstration, a filtering switch operating at 780 MHz is implemented with the compact size of 0.072λg×0.066λg. More than 20-dB suppression levels are measured from 0.89 to 5.3 GHz (6.8 f0), featuring wide-stopband responses.

Underwater optical cameras are widely used for security monitoring in ocean, such as earthquake prediction and tsunami alarming. Optical cameras recognize objects for autonomous underwater vehicles and provide security protection for sea-floor networks. However, there are many issues for underwater optical imaging, such as forward and backward scattering, light absorption, and sea snow. Many underwater image processing techniques have been proposed to overcome these issues. Among these techniques, the depth map gives important information for many applications of the post-processing. In this paper, we propose a Kinect-based underwater depth map estimation method that uses a captured coarse depth map by Kinect with the loss of depth information. To overcome the drawbacks of low accuracy of coarse depth maps, we propose a corresponding reconstruction architecture that uses the underwater dual channels prior dehazing model, weighted enhanced image mode filtering, and inpainting. Our proposed method considers the influence of mud sediments in water and performs better than the traditional methods. The experimental results demonstrated that, after inpainting, dehazing, and interpolation, our proposed method can create high-accuracy depth maps.

With the rapid development of smart phones and wireless technology, mobile services and applications in the world are growing rapidly. Advanced mobile computing and communications greatly enhance users’ experience by the notion of “carrying small while enjoying large”, which have brought a huge impact to all aspects of people’s lifestyles in terms of work, social, and economy. Although these advanced techniques have extensively improved users’ quality of experience (QoE), it is not adequate to provide affective services without efficient mechanisms of emotion-aware mobile computing, which includes various unique aspects, e.g., mobile data sensing and transmissions; sentiment analysis and emotion recognition; affective interaction. Under the new service paradigm, novel mobile services and innovative applications need to be extensively investigated to gain the great potentials brought by emotion-aware mobile computing.

Unstructured clinical medical text, as an important part of the electronic health records, is characterized by large quantities and can store substantial disease-related information of patients. Currently, the disease risk assessment model based on the analysis of clinical medical text designs relevant characteristics aiming at certain diseases, and different characteristics are identified from the text using different methods. In this way, changes of disease performance characteristics are difficult to adapt. Furthermore, it is hard to use the risk assessment model in other disease applications. As a result, this paper establishes the universal disease risk assessment model using the data of clinical medical text, conducts the independent study, and extracts disease characteristics from substantial historical data to avoid the limitations designing disease characteristics. First, this paper analyzes the medial clinical text to determine the contents related to the disease characteristics of patients. Second, learn the representation of clinical text with unsupervised learning methods, and study and extract the disease characteristics from the substantial historical data of patients in the convolutional neural network to assess disease risks. Finally, make a contrast experiment of disease risk assessment using the clinical text data of patients with cerebral infarction, patients with pulmonary infection, and patients with coronary atherosclerotic heart disease from the data of a second grade-A hospital in China and related methods. The experiments show that the approach proposed in this paper achieves the disease risk assessment for different diseases with the same structure.

To detect multiple sclerosis (MS) diseases early, we proposed a novel method on the hardware of magnetic resonance imaging, and on the software of three successful methods: biorthogonal wavelet transform, kernel principal component analysis, and logistic regression. The materials were 676 MR slices containing plaques from 38 MS patients, and 880 MR slices from 34 healthy controls. The statistical analysis showed our method achieved a sensitivity of 97.12±.14%, a specificity of 98.25±0.16%, and an accuracy of 97.76±0.10%. Our method is superior to five state-of-the-art approaches in MS detection.