Search

For a class of permanent magnet synchronous motors (PMSM) characterized by strong coupling, significant nonlinearity, load uncertainties, and external disturbances, this paper investigates an extended state observer-based direct model-free adaptive predictive compensation control (ESO-dMFAPCC) method for speed regulation. Initially, a dynamic linearization model of a PMSM system with unknown external disturbances is developed. Subsequently, a novel direct model-free adaptive predictive compensation controller, theoretically equivalent to an ideal controller, is proposed. An extended state observer (ESO) is designed to estimate the state variable, capturing the characteristics of the external disturbance and system uncertainty. On this basis, control gains with online adaptive capabilities are designed to achieve adaptive speed control for the PMSM system. The primary advantage of the proposed control method lies in its novel framework for directly constructing a predictive compensation controller, which simplifies the controller structure design process by reducing it to a parameter-optimization problem. Moreover, utilizing only the PMSM system’s input/output (I/O) data, the proposed ESO-dMFAPCC is purely data-driven and exhibits strong robustness against external disturbances. Furthermore, the stability of the closed-loop PMSM system is rigorously analyzed. Finally, the simulation results demonstrate that, compared to PI control, sliding mode control (SMC), and model-free adaptive predictive control (MFAPC), the proposed method offers superior dynamic performance while significantly enhancing the anti-disturbance capabilities of the PMSM system.

With the widespread application of mobile robotics technology, path planning has increasingly become a research hotspot. In complex environments, planning an efficient, stable, and safe path is an urgent problem that needs to be addressed. To this end, this paper proposes the Improved Walrus Optimization Algorithm (IWaOA) and applies it to path planning. Firstly, the Sine-Tent-Cosine chaotic mapping is used to initialize the walrus population, addressing the issue of insufficient population diversity in the later stages of the algorithm’s iteration. Next, two improvement strategies are proposed: optimal value-enhanced random walk and directional evolutionary mutation. These strategies aim to enhance the algorithm’s local search capability and precision, optimizing the issues of the original algorithm’s proneness to falling into local optima and slow convergence speed. Finally, building on the three stages of the Walrus Optimization Algorithm (WaOA), this paper introduces a fourth stage termed the “Hunting Stage” to the original algorithm with historical experience positions. It’s capable of significantly improving the overall performance of the algorithm. Evaluating the performance of the proposed algorithm, this paper conducts experiments with three distinct sets of benchmark functions and compares the outcomes against various swarm intelligence algorithms. Furthermore, the IWaOA was applied to the path planning problem for mobile robots. The experimental results confirm the efficacy and advantage of the IWaOA compared to the traditional WaOA, demonstrating a decrease in path length by 16.7%, 3.7%, and 6.2% across three different map scenarios.

This study presents an enhanced deadbeat predictive control (DPC) method for permanent magnet synchronous motor (PMSM) drive system. By taking the model parameter mismatch and harmonics induced by non-sinusoidal back electromotive force (EMF) into consideration, an improved predictive error model is presented. Then, a parallel-observer combining super-twisting sliding model observer (STSMO) and adaptive line neuron (ADALINE) is proposed to estimate the fundamental and sixth harmonic components of predictive error model. Furthermore, an enhanced parallel-observer based DPC method is proposed to design rotor current controller. The advantage is that the predictive error under non-sinusoidal back EMF is reduced and the sixth harmonic in current is suppressed obviously. Better predicting and tracking performance can be guaranteed compared with conventional DPC method. Finally, numerous simulation and experimental results are given to validate the effectiveness and robustness of proposed method. It can be seen from experiment results that the tracking offset is tiny and the sixth harmonic ratio of predictive error is reduced to 0.06%. It indicates that the harmonic suppression performance and suitability of proposed method are satisfactory.

In recent years, constrained by the challenges associated with expensive data annotation and poor generalization ability in supervised models, domain adaptation has been proposed to effectively mitigate domain gaps, which has gained significant attention in the field of remote sensing. However, most existing methods do not consider the inherent characteristics of remote sensing images when formulating adaptive strategies. In addition, there has been limited research on addressing class-imbalanced data situations, leading to undesirable performance in domain adaptation tasks involving long-tailed datasets. To overcome the aforementioned limitations, based on the analysis of intraclass diversity and intradomain style differences of remote sensing images, we propose a novel prototype contrastive learning framework called general-specific prototype contrastive learning (GSPCL). Due to the unreliability of clustering samples in conventional prototype-based clustering methods, the bidirectional weighted prototype strategy is proposed to optimize this loophole. Consequently, more robust prototypes are constructed in both domains, serving as mediators to reduce domain discrepancies bidirectionally. Particularly, often overlooked in most methods, we incorporate low-confidence sample features into the contrastive learning process alongside these prototypes to further guide the model to address feature alignment and long-tail issues effectively. Finally, in order to verify the superiority of our proposed method, we adhere to two existing experimental settings and construct an extra optical remote sensing domain adaptation dataset with class-imbalanced scenarios. In the first two experimental settings, GSPCL outperforms the second-ranked approach by 5.0% and 4.0% in average accuracy. Furthermore, our approach exhibits highly competitive results in handling long-tailed data scenarios.

Phase unwrapping (PU) is one of the core procedures of interferometric synthetic aperture radar. The traditional PU methods estimate the absolute phase gradient based on the assumption that phase is spatially continuous. However, this assumption is often not true due to the large-phase gradient. In this article, a local enhanced U-net (LEU-Net) was developed to determine the phase discontinuity of synthetic aperture radar (SAR) differential interferograms in areas with substantial elevation change. The phase discontinuity in interferograms predicted by this network was used as a priori information for the phase unwrapping maximum flow\minimum cut algorithm (PUMA). Within the LEU-Net, an interactive compression module was used to reduce the loss of detail information and ensure the ability of the model to segment discrete phase discontinuities; and a cross-fusion module was used to fuse multilevel features and highlight the location information while suppressing noise. The phase discontinuity information predicted by the LEU-Net guided the PUMA to obtain the unwrapped phase through iterative optimization; and in the meantime, the information can suppress the propagation of PU errors from areas of low signal-to-noise areas to areas of high quality. Experiments on simulated and real SAR interferograms with large-phase gradients showed that our method has a more robust PU capability.

With the maturity of satellite imaging technology, satellite video has attracted more and more attention because of high spatial resolution and temporal resolution. Thus, object tracking as the main application task of satellite videos also becomes a popular research topic. Compared to natural videos, the difficulties of satellite videos object tracking mainly are caused by feature deficiency of small objectives, surroundings, occlusion (OCC), etc. In this article, based on a dual correlation filter (DCF) tracking framework, a new object tracking method for satellite videos is proposed to deal with the above-mentioned problems. For feature deficiency problem, the proposed super-resolution feature enhancement module could improve the feature discriminative ability of objects utilizing the prior information learned by a super-resolution network. For the OCC problem, a multilayer perceptron motion prediction network is designed to predict the position of objects when OCC occurs, which uses an online training strategy. Besides, KNN background subtraction also is introduced to reduce the interference of surroundings. Finally, these above-mentioned processes are combined together appropriately on the DCF tracking framework for better tracking results. Aiming to verify the performance of the proposed method, abundant experiments have been conducted on satellite video datasets. The experimental results show that the design of the proposed tracking method is effective and it also has a conspicuous advantage compared with some state-of-the-art methods.

In the 5G era, more sensitive information will be transmitted on optical networks. Although modern cryptography could support different security mechanisms in different layers, the optical layer has not been paid enough attention but it is also vulnerable to attack by tapping or other methods. The optoelectronic firewall is one of the promising security strategies which can accelerate the operating speed within manageable costs by employing optical signal processing compared with current electronic firewalls. The most significant and challenging component of an optoelectronic firewall is pattern recognition. However, either the latest pattern recognition systems can only process the OOK or BPSK modulation format signals, or the coherent correlators supporting higher order modulation format require advanced components to deal with a higher number of symbols. In this paper, to address the pattern recognition challenges, we propose, analyze and simulate a pattern recognition system of QPSK signals integrated into all-optical high-speed optoelectronic firewalls used for optical layer security. The numerical simulation results first demonstrate the feasibility, and a baud rate of 100GBaud can be achieved. Next, a threshold-setting method is developed which also proves the ability to recognize arbitrary target patterns. Then some noise analysis results demonstrate that the pattern recognition of the I branch can be immune to the noise while the output power of EDFAs in the Q branch needs to be configured carefully. Finally, the matched result of a real Ethernet frame further reveals that our proposed system is promisingly applied in optoelectronic firewalls.

Transcranial direct current stimulation (tDCS) has become a new method of rehabilitation treatment for stroke and is gradually gaining acceptance. However, the neurophysiological mechanisms of tDCS in the treatment of stroke still need to be further explored. In our research, we aimed to investigate the effects of tDCS on neural oscillation power and brain functional network (BFN) connectivity in stroke patients based on electroencephalogram (EEG). Fifteen patients with ischemic stroke attended two experimental sessions in a randomized crossover trial and received real and sham tDCS. Resting-state EEG signals were acquired before and after stimulation. First, the power of EEG delta, theta, alpha, beta, and gamma bands were analyzed before and after stimulation. Then, BFNs in different bands were constructed, and network characteristic parameters such as degree, clustering coefficient, characteristic path length, global efficiency, and small-world attribute were analyzed. The results showed that the delta power decreased significantly ( ${P}< {0.05}$ ) and the alpha power increased significantly ( ${P}< {0.05}$ ) after real tDCS, whereas the theta, beta, and gamma power did not change significantly ( ${P}>{0.05}$ ). The degree and global efficiency of the delta band BFN decreased significantly after real tDCS ( ${P}< {0.05}$ ), the characteristic path length increased significantly ( ${P}< {0.05}$ ), and the clustering coefficient and small-world attribute decreased but not statistically significant ( ${P}>{0.05}$ ). The degree, clustering coefficient, global efficiency, and small-world attribute of the alpha band BFN increased significantly after real tDCS ( ${P}< {0.05}$ ), and the characteristic path length decreased significantly ( ${P}< {0.05}$ ). The BFN characteristic parameters in theta, beta, and gamma bands did not change significantly after real tDCS ( ${P}>{0.05}$ ). There was no significant difference in sham tDCS group ( ${P}>{0.05}$ ). Our study found that the power of delta oscillations decreased and the global connectivity of delta band BFN weakened after tDCS in stroke patients, whereas the power of alpha oscillations increased and the global and local connectivity of alpha band BFN was enhanced. The findings implied that one of the neurophysiological mechanisms of tDCS treatment in stroke patients may be the modulation of the power of delta and alpha oscillations, as well as the improvement of the connectivity of delta and alpha oscillation networks.

State estimation has long been a fundamental problem in signal processing and control areas. The main challenge is to design filters with ability to reject or attenuate various disturbances. With the arrival of Big Data era, the disturbances of complicated systems are physically multisource, mathematically heterogenous, affecting the system dynamics via isomeric (additive, multiplicative, and recessive) channels, and deeply coupled with each other. In traditional filtering schemes, the multisource heterogenous disturbances are usually simplified as a lumped one so that the “single” disturbance can be either rejected or attenuated. Since the pioneering work in 2012 (Guo and Cao, 2012), a novel state estimation methodology called composite disturbance filtering (CDF) has been proposed, which deals with the multisource, heterogenous, and isomeric disturbances based on their specific characteristics. With CDF, enhanced antidisturbance capability can be achieved via refined quantification, effective separation, and simultaneous rejection and attenuation of the disturbances. In this article, an overview of the CDF scheme is provided, which includes the basic principle, general design procedure, application scenarios (e.g., alignment, localization, and navigation), and future research directions. In summary, it is expected that CDF offers an effective tool for state estimation, especially in the presence of multisource heterogeneous disturbances.

Due to the great uncertainty of the output power of renewable energy sources (RESs), the storage and reuse of RESs is the focus of new energy technology. The bidirectional DC-DC resonant converter is preferred in the new energy technology. However, the efficiency of traditional bidirectional DC-DC resonant converter will be greatly affected by the wide voltage gain range. By applying the adaptive excitation inductance, a CLLCLC resonant converter is proposed. The $L-C$ branch is introduced to replace the fixed excitation inductance in traditional CLLLC resonant converter. In the design of transformer, the turns ratio and excitation inductance of transformer are decoupled to reduce the difficulty of transformer design. Simultaneously, compared with the traditional transformer, the impedance of equivalent excitation inductance will adjust automatically against the working frequency, decreasing the circulating current loss of resonant tank. Therefore, the proposed converter solves the contradiction between wide voltage gain range and high efficiency existed in the traditional converter. Finally, an experimental platform with 200 W rated power is built to verify the effectiveness of proposed topology.

This paper proposes a non-isolated three-port converter which integrates the input port, battery port, and load port into one converter for renewable energy applications. The coupled inductor and switched capacitor are used to achieve high voltage gain and three power switches can be utilized to realize the power flows between the sources, the battery, and the load. The energy stored in the leakage inductance is recycled to reduce the voltage stress of the power switch. In addition, various operating stages are analyzed and design considerations are presented. Compared to the relevant converters, the proposed converter can realize power flows and achieve high voltage gain by using few components. Finally, a laboratory prototype of the proposed converter with input port voltage 24 V, battery port voltage 48 V, and output voltage 400 V with 200 W rated power is implemented to validate the feasibility and effectiveness of the theoretical analyses.

Catastrophic forgetting is a key challenge for class-incremental learning with deep neural networks, where the performance decreases considerably while dealing with long sequences of new classes. To tackle this issue, in this paper, we propose a new exemplar-supported representation for incremental learning (ESRIL) approach that consists of three components. First, we use memory aware synapses (MAS) pre-trained on the ImageNet to retain the ability of robust representation learning and classification for old classes from the perspective of the model. Second, exemplar-based subspace clustering (ESC) is utilized to construct the exemplar set, which can keep the performance from various views of the data. Third, the nearest class multiple centroids (NCMC) is used as the classifier to save the training cost of the fully connected layer of MAS when the criterion is met. Intensive experiments and analyses are presented to show the influence of various backbone structures and the effectiveness of different components in our model. Experiments on several general-purpose and fine-grained image recognition datasets have fully demonstrated the efficacy of the proposed methodology.

Task-based functional magnetic resonance imaging (tfMRI) is a widely used neuroimaging technique in exploring brain networks and functions associated with cognitive behaviors. Traditionally, the general linear model (GLM) is the most popular method in tfMRI data analysis due to its simpleness and robustness. This model-driven method adopts a canonical hemodynamic response function (HRF) and its various derivatives to construct regressors in the design matrix and estimate changes in the tfMRI data. However, a possible limitation of current model-driven methods is that the HRF is fixed and non-adaptive which may overlook other diverse and concurrent brain networks. In order to overcome these limitations, we proposed a novel hybrid framework, supervised brain network learning based on deep recurrent neural networks (SUDRNN), to reconstruct the diverse and concurrent functional brain networks. Specifically, this hybrid framework first takes advantage of the great capacity of deep recurrent neural networks (DRNN) in modeling sequential data to learn the diverse regressors from real tfMRI data. After that, it utilizes the effective supervised dictionary learning (SDL) method to reconstruct both the task-related functional brain networks and other latent brain networks simultaneously. Extensive experiment results on different tfMRI datasets from Human connectome project (HCP) demonstrated the superiority of the proposed framework.

Transcranial direct current stimulation (tDCS) is an emerging brain intervention technique. However, the effects of tDCS on stroke patients are unclear. In this study, Granger Causality (GC) is used to construct brain functional networks of the tDCS group (9 cases) and the sham tDCS group (11 cases). Some parameters of brain functional networks are analyzed comparatively, such as degree, clustering coefficient, betweenness centrality, characteristic path length, and efficiency. In tDCS group, the results show that the distribution of degree changes greatly in left prefrontal lobe and left temporal lobe regions, the distribution of betweenness centrality transfers, characteristic path length decreases, clustering coefficient and efficiency increase significantly (p

The biological brain has self-adaptive ability through neural information processing and regulation. Drawing from the advantage of the biological brain, it is significant to research the robustness of artificial neural network (ANN) based on brain-like intelligence. In this study, based on the Izhikevich neuron model and the synaptic plasticity model which contains excitatory and inhibitory synapses, a spiking neural network (SNN) with small-world topology and a SNN with scale-free topology are constructed. The anti-injury function of two complex SNNs (CSNN) under random attack is comparatively analyzed. On this basis, the information processing of CSNN under attack is further discussed, and the anti-injury mechanism of CSNN is explored based on the synaptic plasticity. The experimental results show that: (1) scale-free SNN (SFSNN) has better performance than small-world SNN (SWSNN) in the anti-injury ability under random attacks. (2) The information processing of CSNN under random attacks is clarified by the linkage effects of dynamic changes in neuron firing, synaptic weight, and topological characteristics. (3) The anti-injury ability of CSNNs is closely related to the dynamic evolution of synaptic weight, which implies the dynamic regulation of synaptic plasticity is the intrinsic factor of the anti-injury function of CSNNs. This study lays a theoretical foundation for the application of brain-like intelligence with adaptive fault-tolerance.

Remote sensing image scene classification, which aims at labeling remote sensing images with a set of semantic categories based on their contents, has broad applications in a range of fields. Propelled by the powerful feature learning capabilities of deep neural networks, remote sensing image scene classification driven by deep learning has drawn remarkable attention and achieved significant breakthroughs. However, to the best of our knowledge, a comprehensive review of recent achievements regarding deep learning for scene classification of remote sensing images is still lacking. Considering the rapid evolution of this field, this article provides a systematic survey of deep learning methods for remote sensing image scene classification by covering more than 160 papers. To be specific, we discuss the main challenges of remote sensing image scene classification and survey: first, autoencoder-based remote sensing image scene classification methods; second, convolutional neural network-based remote sensing image scene classification methods; and third, generative adversarial network-based remote sensing image scene classification methods. In addition, we introduce the benchmarks used for remote sensing image scene classification and summarize the performance of more than two dozen of representative algorithms on three commonly used benchmark datasets. Finally, we discuss the promising opportunities for further research.

The migratory animals are capable of traveling long distances by fusing varied information, such as the halteres, the star positions, and the polarization pattern. Motivated by the navigation strategy of migratory animals, a bio-inspired navigation strategy based on polarized skylight, starlight, and halteres is presented in this paper. To enhance the environmental adaptability of the polarization navigation system, a sun vector calculation method tightly fusing the degree of polarization (DoP) and the azimuth of polarization (e-vector) is proposed. By setting the threshold of DoP values of the observation points, the DoP and the e-vector within the threshold range can be selected to deduce the sun position. As such, the robustness and reliability of the sun position can be improved. Concerned with the issue of attitude and heading determination in GNSS-challenged conditions, an integrated navigation strategy combined with polarized light and starlight is proposed. In addition, a Kalman filter is adopted to fuse the navigation data of polarized light, starlight, and inertial sensors. Finally, simulations are conducted to validate the performance of the integrated model. The results illustrated that the proposed navigation strategy is capable of determining the attitude and heading in the scenarios of which the Global Navigation Satellite System (GNSS) is disturbed.

As a part of the energy transmission chain, gearboxes are considered as important components in rotating machines, and the gearbox failure results in costly economic losses. Therefore, it is necessary to detect the appearance of incipient gearbox faults by implementing an appropriate detected model. The incipient failure characteristics of the gearbox are weak and hidden in a set of time-varying series signals the vibration signals, which is difficult to effectively extract under the background of strong noise. The PCA method is not effective in detecting weak fault features in time-varying signals, so this paper proposes a method based on Deep Recursive Dynamic Principal Component Analysis (Deep RDPCA) to detect incipient faults in gearboxes. The proposed approach is modeled via both the deep decomposed theorems and time-varying dynamic model based on traditional PCA to extract characteristic of time-varying and weak fault information under the background of strong noise. The proposed method could get a better real-time reflection for changed system by introducing “Moving Window” technologies, so that the incipient fault of gearbox could be detected accurately, too. Finally, the effect of Deep RDPCA-based fault diagnosis is compared with the results of PCA, DPCA, RDPCA, Deep PCA, and Deep DPCA methods. It is concluded that the proposed method can effectively capture the time-varying relationship of process variables and accurately extract the weak fault characteristics in the vibration signal, which effectively improves the fault detection performance.

In the study of image classification, neural network learning relies heavily on datasets. Due to variability in the difficulty of collecting images in reality, datasets tend to have class imbalance problems, which undoubtedly increases the difficulty of classification. During the training of a neural network, classes with a large number of images are naturally trained more often than classes with a small number of images. Because of imbalanced training, the classification ability of neural networks on test and validation sets differs greatly in different categories. The test results of more training classes are better, and the test results of classes with less training are poor. In this paper, we propose two kinds of balanced loss functions, namely, CEFL loss and CEFL2 loss, by rebalancing the cross-entropy loss function and focal loss function. The experimental results show that the proposed loss functions are significantly able to improve classification accuracy on class-imbalanced datasets.

Studying the moisture diffusion in non-oil-immersed insulation paper is helpful to optimize the drying process and improve the drying efficiency of oil-immersed equipment. Moisture diffusion factor (MDF) is the key parameter to describe the moisture diffusion in non-oil-immersed insulation paper. Thermal aging affects the moisture absorption capacity of insulation paper obviously, thus, this paper intends to modify the empirical expression of MDF by considering the effects of thermal aging. The transient moisture contents of non-oil-immersed insulation paper with different thicknesses were measured at 40 °C and 60 °C, respectively. Then, the experimental values of MDF were extracted, and the empirical expression of MDF was modified by considering the effects of thermal aging. Finally, the modified expression of MDF was verified. The following conclusion can be obtained: 1) the equilibrium time constant of non-oil-immersed insulation paper increases with thermal aging; 2) the moisture absorption rate of non-oil-immersed insulation paper decreases with thermal aging, and; 3) the modified expression of MDF is reasonable and effective.

Internet of Things (IoT) plays as the key infrastructure for various applications, e.g., smart city, smart grid, and smart manufacturing. Wireless communication is greatly limited for IoT devices due to the severe electromagnetic interference in industrial radio-hostile environments. Visible light communication (VLC) has been considered as a promising communication technique that could complement the wireless communication. Due to complex industrial environments, it is not practical to wire such scale of VLC-enabled IoT devices into the power line or replace batteries for them. To cope with this challenge, we propose to harvest the energy from electromagnetic interference signals to power the VLC for IoT devices. We achieve this goal by proposing advanced signal processing designs, new energy-efficient schemes, and seamless integration. Finally, we prototype the system on real devices, and our extensive experiments show that our system can achieve promising results.

The present “condition-based maintenance decision making” of power transformers will cause large financial losses to electric enterprises, because of not having taken reliability and economy into consideration. To solve this problem, a maintenance decision-making model with the consideration of reliability and economy was established to choose the best maintenance strategy for oil-filled transformers. With the corrected parameters of operating environment and maintenance records, a condition assessment model including DGA test, oil test, and the electrical test was proposed to decide the comprehensive health index of transformers. After establishing the relationship between the fault rate and the comprehensive health index, a reliability evaluation model was formed, which can simulate the impact of different maintenance types. Taking the reliability and economy operation of transformers into consideration, a particle swarm optimization method was developed to solve the optimization model and select the best maintenance strategy according to the current condition of transformers. Two cases were studied and the results demonstrate that the proposed model offers an improved maintenance strategy.

Sapphire is one of the most widely used electronic substrate materials in the industry. The manufacturing process of sapphire and such hard and brittle materials is extremely time and energy consuming because of their superior material properties in physical and chemical behaviors. In this study, the ultraviolet-curable resin is introduced into the fabrication of the diamond abrasive lapping plate as a bonding agent. A practical manufacturing process is established in the laboratory to develop an ultraviolet-curable resin bond fixed abrasive lapping plate for the precision machining of sapphire substrates, the machining performance of which is examined through a series of comparative experiments among the conventional fixed-abrasive lapping, the slurry-based lapping and this originally developed plate involved lapping. The surface topography and surface roughness of the machined sapphire workpieces are measured to evaluate the surface quality, and the weight loss of the workpieces in each machining process is recorded to estimate the respective machining efficiency. The promising results achieved from the ultraviolet-curable resin plate lapping process, in terms of a relatively higher material removal rate and better surface quality, indicate that the occurrence of a corporate action integrating both the advantages of the fixed abrasive grains and loose abrasive grains. A summarized hypothesis is drawn to describe the dynamically balanced state of the hybrid precision abrasive machining process, in which the 2-body abrasion and 3-body abrasion material removal mechanism participate simultaneously and interconvert to each other continuously.

Stepped impedance resonator (SIR) has been widely used for applications in wide stopband bandpass filters. However, the reported SIR wide stopband bandpass filters are limited in either stopband performance or design complexity. Here we propose an approach to improve the stopband performance and reduce the design complexity simultaneously. Two in-resonator open stubs are adopted in a SIR (with tap-feeding) to stagger the frequencies of higher-order modes (compared with other SIR) and produce two adjustable transmission zeros without affecting the fundamental properties. For a SIR filter employing such a structure, the spurious passbands are doubly deteriorated by the staggering and up to four transmission zeros. At the same time, the fundamental passband is not affected, which significantly reduces the design complexity. As a result, not only wide stopband but also high-suppression stopband can be obtained in a simple design. A prototype based on the substrate integrated coaxial line (SICL) is experimented showing an impressive stopband performance. Agreeing with the simulation well, the measured stopband is extended up to about 12 f0 with the suppression of 40 - 50 dB (f0 is the center frequency). The proposed technique could be very useful to design a bandpass filter with a wide and high-suppression stopband efficiently.

To deal with the inherent data sparsity and cold-start problem, many recommender systems try to exploit the textual information for improving prediction accuracy. Due to the significant progress of deep learning techniques, neural network-based content modeling methods have been investigated in recent studies. However, most of these existing methods often assume that users are independent and identically distributed (i.i.d), and the social influence is not considered. However, in the real world, we always turn to our friends for recommendations, and the closer the friendship between the friends, the greater the social impact is. These methods only exploit the reviews from an item perspective and rarely consider the user's reviews to capture the user's interests, but in reality, users often express their preferences by posting different reviews to different items. Based on the above-mentioned observations, we propose a social-enhanced content-aware recommendation method by fusing the social network, item's reviews, and user's reviews in a unified framework. Specifically, to better model the item's reviews, we first introduce the convolutional matrix factorization (ConvMF) as our basic recommendation framework, which utilizes convolutional neural network (CNN) to capture the deeper understanding of the content context. Then, to consider the user's social influence, we integrate the user's social network into ConvMF by a shared user latent factor, which can bridge the user's social interests and user's general preferences in the same latent space. To model the user's reviews, similar to ConvMF, we exploit another CNN to learn a deeper understanding of the user's posted contents. Finally, we conduct experiments on the real-world dataset Yelp to demonstrate the effectiveness of our method. The experimental results indicate that our proposed method can effectively model the social and the review information and outperforms other related methods in terms of root mean squared error (RMSE) and mean absolute error (MAE).

Oil-immersed insulation paper (OIP) bushing is one of the most important equipment in power transmission and distribution systems. Moisture is one of the most serious factors that affect the insulation status of OIP bushing. In this paper, the OIP bushing was assumed to be in a stable and simplicity condition, and then the transient moisture distribution of capacitor core and its effects on the electric field were studied. First, by dissecting a 110 kV OIP bushing, the specific sizes for modelling were obtained. Afterwards, a moisture diffusion model for OIP bushing was established based on COMSOL Multiphysics software, and it was also verified by moisture diffusion experiment. Then, the moisture distribution for different conditions were obtained and analyzed. Finally, a model for the electric field of OIP bushing was established, and then the effects of moisture on the electric field were studied.

The outliers caused by noise and mismatching severely restrict the precision of visual odometry. Moreover, the dynamic environment is also a crucial element that decreases the robustness of the systems. This paper presents a robust stereo visual odometry by decoupled ego-motion estimation based on probabilistic matches and rejecting the outliers of dynamic objects through motion segmentation. Fast ZNCC method, based on local sum table and partition upper bound schemes, is presented for selecting probabilistic matches while keeping run-time efficiency. The selection of multi-correspondences can avoid mismatching of corresponding points. In consideration of noise interference, the essential matrix is computed in a probabilistic framework to estimate the initial value of the rotation matrix without estimated depth errors involved. Then, in order to estimate pose robustly in dynamic environment, a modified sparse subspace clustering (SSC) method is discussed, which aims to cluster the tracked 3D points cloud to avoid errors caused by affine transformation. The non-negative constraint makes the method suitable for fast moving camera. The proposed 3D-SSC method removes the outliers belonging to dynamic objects effectively. Finally, the detected inliers and depths are employed to estimate the translation matrix and refine rotation matrix. The proposed method is evaluated on the KITTI benchmark and compared with the state-of-the-art methods. The results show that our method is more robust as it can detect outliers more accurately in dynamic environments and achieve higher precision in motion estimation.

The blooming of the Internet of Things (IoT) needs a communication infrastructure to allow connecting a wide range of devices. Visible light communication (VLC) provides an attractive solution for these IoT devices. However, the broadcast nature makes the VLC prone to be exposed in the reach of eavesdroppers, which undesirably impairs the IoT system security. The negative factors in practical use, such as light reflection, light intensity, channel correlation, and channel estimation error, remain less touched. In this paper, we propose the SecLight to support both the multiple-input single-output VLC (MISO-VLC) and the single-input single-output VLC (SISO-VLC) IoT devices. A random time reversal scheme is first proposed to automatically steer the SISO-VLC signal to the legitimate receivers by scrambling the eavesdropper's channel. Then, we mathematically analyze the impact of channel correlation and channel estimation error inherent in MISO-VLC system and propose an effective approach to alleviate the decreased secrecy capacity by introducing the Karhunen-LoEve transform and equivalent artificial noise. Through extensive evaluation, the proposed SecLight framework proves to be able to aggravate the bit error rate of eavesdroppers, even when a malicious interception is intended while holding the secrecy capacity above 3 bit/s/Hz. Overall, our results demonstrate that our framework SecLight is a promising solution for enhancing the security of the IoT devices.

In response to the rapid change of attitude in complex situations such as sharp turns and camera jitter, visual odometry/visual simultaneous localization and mapping (VO/VSLAM) often requires a high camera frame rate. However, the high frame rate of the camera poses a great challenge to real-time VO/VSLAM, especially when the vehicle travels along a relatively flat trajectory which requires higher real-time performance. Therefore, in this paper, we propose an automatic method for key-frame selection based on the motion state of the vehicle, which can reduce data redundancy and improve the real-time performance and robustness of VO/VSLAM. First, a pyramid-layered Kanade-Lucas-Tomasi (KLT) algorithm is used to track the feature points, and the five-point method and RANSAC are used to calculate the essential matrix. Second, two-step decomposition is used to calculate the change of inter-frame attitude, and then, key-frames are automatically selected based on the motion state of the vehicle predicted by attitude change within a short time interval. To evaluate the method, we conduct extensive experiments on real data and Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI) dataset based on monocular visual odometry. Then, we made qualitative and quantitative evaluations from the comparison with the reference trajectory, relative error (RE), root mean square error (RMSE), and absolute trajectory error (ATE). The results indicate that our method can improve the real-time performance while ensuring accuracy, with the data redundancy reduced by about 40%-60%. In addition, the performance of our method is further verified by comparison with the current representative ORB-SLAM.

The complete failure of the rolling bearing is a deterioration process from the initial minor fault to the serious fault, it is meaningless for guiding maintenance when the serious fault is alarmed. This work presents a novel initial fault diagnosis framework based on sliding window stacked denoising auto-encoder (SDAE) and long short-term memory (LSTM) model. In this approach, multiple vibration value of the rolling bearings are entered into SDAE by sliding window processing. Then, multiple vibration value of the rolling bearings of the next period is predicted from the signal reconstructed by the trained SDAE in the previous period using LSTM. For the given input data, the reconstruction errors between the next period data and the output data generated by trained LSTM are used to detect initial anomalous conditions. The proposed method not only utilizes the ability of SDAE to learn the inherent distribution of data, but also ensures that LSTM can extract timing relationships between data cycles, and the model is built using only normal data. The initial fault detection as a key difficulty in the operating condition monitoring and performance degradation assessment of the rolling bearing is effectively solved. Experimental and classic rotating machinery datasets have been employed to testify the effectiveness of the proposed method and its preponderance over some state-of-the-art methods. The experiment results indicate that the proposed method can effectively detect the initial anomalies of the rolling bearing and accurately describe the deterioration trend with strong robustness, and have high significance for maintenance guiding.

In pattern matching, a gap constraint is a more flexible wildcard than traditional wildcards “?”and “*”. Pattern matching with gap constraints is more difficult to handle and fulfills user's enquiries more easily. Pattern matching with gap constraints has therefore been carried out in numerous research works, such as music information retrieval, searching protein sites, and sequence pattern mining. Strict pattern matching under a nonoverlapping condition, as a type of pattern matching with gap constraints, is a key issue of sequence pattern mining with gap constraints since it can be used to compute the frequency of a pattern. Exact matching limits the flexibility of the match to some extent since it requires each character to be matched exactly. We therefore address approximate strict pattern matching under the nonoverlapping constraints (ASPNO) and propose an effective algorithm, named NETtree for ASPNO (NETASPNO), which first transforms the problem into a Nettree data structure, an extensive tree structure. To find the nonoverlapping occurrences effectively, we propose the concept of number of roots paths with distance constraints (NRPDC) which indicates the number of path from a node to the roots with distanced and can be used to delete useless parent-child relationships and useless nodes. We iteratively recalculate the NRPDCs of each node on the subnettree with the rightmost root. Then we can get a path from the rightmost leaf to its rightmost root without using the backtracking strategy. NETASPNO therefore iteratively gets the rightmost root-leaf-path and prunes the path on the Nettree. Extensive experimental results demonstrate that NETASPNO has better performance than the other competitive algorithms.

Outdoor communication wiring is mainly used to provide a physical connection between telecommunication services. With many fibers to the X networks having been built to provide high-level reliability services and to reduce operational costs, the operator has a strong demand for an automatic optical distribution frame. Based on the mathematics of knots to study the relationship of the fiber-optic connections on the main distribution frame, we proposed the ``HIGHER ALWAYS ON TOP”method for fiber-optic switching. Two prototypes of a 1-D array of input lines and a 1-D array of output lines use a line match line and a line match circle to demonstrate these concepts. We proposed the method of constructing non-coplanar lines in a 1-D array of input lines and a 2-D array of output lines, as well as a 2-D array of input lines and a 2-D array of output lines, for scalability. We imitated the opening of a shuttle in the textile field and proposed a high-speed cross-connect method and a simulation experiment to show its feasibility.

This paper presents the first antenna-integrated socket that can be used in place of the commercial socket. The socket panel is made of K9 glass for the aesthetic purpose. It serves as the dielectric resonator (DR) loading for the integrated slot antenna fabricated in the ground plane. The socket (DR)- loaded slot antenna is directly fed by a coaxial cable at the slot center. Considering the practical scenarios, the antenna is investigated in three different cases, namely the socket panel, socket panel with plug, and socket panel with the plug and cable. For demonstration, the socket-loaded antenna is designed in 2.4-GHz WLAN band (2.40–2.48 GHz). The reflection coefficient, radiation pattern, antenna gain, and efficiency in each case are measured. Measured results show that the impedance bandwidth can cover the entire frequency band. The antenna can be potentially used as a camouflaged antenna or an attractive home or office decoration.

With the prosperity of wireless networks, vehicular networks (VNs) have been extensively studied in recent years. It is deployed to ensure road safety, enhance the driving experience, and reduce traffic congestion. However, VNs are vulnerable to various attacks, especially Distributed Denial of Service (DDoS) that attackers control a large number of compromise nodes inside the networks to occupy the network resources of legitimate users and impact the communication among vehicles and between vehicles and infrastructure. In this paper, we design a platform to efficiently detect and rapidly respond to the DDoS attack in VNs based on software-defined networking (SDN). The proposed platform not only contains the trigger mechanism based on the message of OpenFlow protocol (i.e., PACKET_IN message) for a response not timely but also involves a flow feature extraction strategy based on the multi-dimensional information. Moreover, we construct an effective global network flow table feature values based on OpenFlow flow table feature and the entropy feature of flow table entry. We determine all flow table entry by the trained SVM. By analyzing the simulation results, we verify that the detection scheme effectively reduces the time for starting attack detection and classification recognition and has a lower false alarm rate.

Membrane proteins occupy an important position in the life activities of humans and other species. The elucidation of membrane protein types provides clues for understanding the structure and function of proteins. With the fusion of various protein information including amino acid classification, physicochemical property, and evolutionary information, this paper proposes a system for predicting membrane protein types. In this system, a new feature selection method called MIC-GA is proposed to deal with the curse of high-dimensional features. The findings show that this approach is effective in reducing feature dimensions and improves prediction accuracy. Ensemble method based on stacked generalization is also used to solve the problem of feature heterogeneity. The performance of the present method is evaluated on two benchmark datasets. The overall prediction accuracies of eight types are 89.23% and 93.49% using jackknife test and independent test, respectively. The final experimental results show that our method is more effective than the existing methods for prediction of membrane protein types.

With the advent of complex computing applications such as cloud computing and artificial intelligence, the utilization of multicore processors has become one of the best solutions to improve the computation performance. Benefitted from the technologies of silicon-photonic-based communication and three-dimensional (3-D) integration, the 3-D optical network-on-chip (3-D ONoC) has attained extensive attentions as a new multicore architecture, providing high communication bandwidth, as well as low transmission delay and power consumption. As the main part of 3-D ONoCs, the structures of topology and optical router (OR) heavily affect the transmission efficiency of the whole network. In this paper, we propose a mesh-based topology and a novel cost-efficient 6 × 6 nonblocking OR structure. Different from the traditional 3-D ONoC topology that needs seven-port ORs to realize the data transmission, the OR mentioned in our solutions only consumes six ports. This improvement effectively reduces the number of optical switching elements and cross waveguides in ORs, lowering the power consumption and hardware costs of the system. The simulation analysis based on the modified Noxim simulator demonstrates that our method performs well in terms of mitigating the OR- and network-level insertion loss, shrinking the floorplan/chip area and improving the network scalability over benchmarks.

Intrusion detection system (IDS) is emerging as a technology to improve the security of vehicular ad hoc network. However, the recent research on performance of IDS indicates that monitoring overhead is still a significant issue and challenge. The fundamental problem is caused by the fact that each node repeatedly detects adjacent nodes during operation. At the same time, monitoring results are sent to other nodes in the form of control packets, resulting in the decline of the network throughput. In this paper, we focus on selecting a bunch of monitoring nodes and propose a cross-layer security monitoring selection algorithm based on traffic prediction (CLSM-TP). Instead of repeatedly detection for each node, we select the monitoring node whose idle degree is relatively high by predicting the node's traffic based on a cross-layer vehicular ad hoc network. Moreover, the proposed algorithm leverages the mutual information and residual energy to optimize the node selection through social network analysis. The noteworthy contributions are that CLSM-TP can balance the resource consumption among all nodes and prolong the lifetime of vehicular ad hoc network to some extent. Our experimental results show that, the monitoring nodes selected by the algorithm proposed in this paper with higher idle degree preform good enough to monitor the whole vehicular ad hoc network.

Polarization navigation has been an active research direction in navigation fields due to its autonomy and robustness, especially in global navigation satellite system-denied environment. Most existing bionic polarization sensing schemes employed the linear polarizer as the analyzer. However, the quadrature error between the two sets of mutually orthogonal polarizers is usually inevitable, which may cause the sensor output distortion and directly affect the performance of polarization navigation sensors. In this paper, inspired by the special polarization-sensitive structure of insect compound eyes, a polarizing beam splitter (PBS) based polarization navigation sensor is designed. Concerned with the optical path coupling and extinction ratio inconsistency problem between the transmitted beam and the reflected one of the PBS, the novel sensor model is established, where we introduce a coupling coefficient to solve the problems mentioned above. An algorithm of computing the polarization degree and the polarization azimuth angle is presented based on the model. The unscented Kalman filter calibration method is adopted and the calibration experiments are carried out. Finally, the outdoor static accuracy of the sensor is tested, where the results showed that the sensor designed in this paper achieved high accuracy and good stability for a long time.

Nowadays, vehicular applications for traffic safety, traffic efficiency, and entertainment have put forward higher requirements to vehicular communication systems. The fifth-generation (5G) cellular networks are potential to provide high-capacity low-latency communications for vehicles in highly mobile environments. In addition, the IEEE 802.11p-based dedicated short-range communication technology is proposed to form vehicular ad hoc networks (VANETs). Integrating cluster-based VANETs with 5G cellular networks is beneficial for saving scarce spectrum resources, preventing network congestion, and reducing packet loss. However, it is a challenging problem to find an effective clustering algorithm that has high stability adapting to dynamic VANETs. Considering the advantages of software-defined networking (SDN), in this paper, we present an SDN-enabled social-aware clustering algorithm in the 5G-VANET system, which exploits a social pattern (i.e., vehicles' future routes) prediction model in order to enhance the stability of clusters. Each vehicle's movement is modeled as a discrete time-homogeneous semi-Markov model, where the state transition probability and sojourn time probability distribution are inputs and each vehicle's social pattern is output. The predicted social patterns are subsequently used to create clusters so that vehicles in the same cluster tend to share the same routes. Cluster heads are selected based on the metrics of intervehicle distance, relative speed, and vehicle attributes. We evaluate our algorithm and the results show that it performs better in terms of cluster lifetime and clustering overhead compared with traditional clustering algorithms.

An integrated ambient kinetic and microwave energy scavenger is presented in connection with its strategy of two-level hybridization, namely structural and functional integration of the two dissimilar energy harvesting techniques. Specifically, they are related to two-resonator/transducer integration and simultaneous/collaborative rectification. An F-shaped radiofrequency antenna serves as the microwave energy collector and a Michael Faraday's generator mounted on a mechanical beam cut out of the printed circuit board acts as the electrical current producer. Both microwave and mechanical resonators are optimally designed and accommodated on the same substrate with a compact size comparable to that of a credit card (80 mm × 50 mm). The demonstrated hybrid power harvester has a capacity to operate properly with either microwave or mechanical excitation. When both power sources are accessible, it can provide a notable efficiency enhancement compared with the separate harvesting of two single sources. The measurements show that at a typical injecting AC power of -38 dBm, the rectified DC output power could be increased by at least 50% compared with a direct summation of DC output power through two separate modes. An efficiency enhancement of 85% can be achieved when diode's injecting power sources are -45 dBm. With a strong resilience, such a hybrid power harvester is able to potentially find its wide practical applications.

A delay/disruption tolerant network (DTN) architecture where a “store-carry-forward”strategy is adopted for data transmissions can be utilized in vehicular ad hoc networks (VANETs). The key point of routing in DTN-enabled VANETs is to choose the best node and determine the best time to forward messages. Time-space graph models provide an idea of converting the dynamic routing problems into static ones in deterministic DTNs. But it is a challenge to predict vehicles' future positions in order to obtain the time-space graph. In this paper, to achieve the cost-efficient and reliable routing in DTN-enabled VANETs, a novel timeliness-aware trajectory data mining algorithm is proposed to predict nodes' future positions. A sparse time-space graph is then obtained, based on which, two routing heuristics are proposed. Simulation results demonstrate that our proposed routing algorithms ensure low cost and high reliability over time.

Social users can be flexibly covered by a vehicle traveling along a certain trajectory in a city. This scenario provides opportunities for marketers to utilize vehicles as recommenders targeting potential customers. As a result, the vehicle social network (VSN) concept emerges. Since a marketer usually operates from a remote central office (CO), an effective communication system is urgently required to support data transmission between VSNs and COs. Orthogonal frequency division multiplexing (OFDM) is a modulation technique characterized by high spectral efficiency and dispersion resistance, and intensity modulation and direct detection (IM/DD) is an effective solution for optical OFDM (OOFDM) data transmission. Performing discrete Fourier transform (DFT) and discrete Hartley transform (DHT) operations can enable the satisfaction of various transmission requirements between a VSN and a CO. In this paper, we support the above claims through theoretical and simulation-based analyses. Moreover, a software-reconfigurable platform is established to achieve flexibility between DFT and DHT functionalities. Thus, diverse transmission requirements can be satisfied by modifying individual modules deployed in a single smart system instead of using multiple rigid network architectures. Simulated and experimental results both demonstrate the feasibility and effectiveness of our solutions.

Survivability in the OpenFlow-based software-defined elastic optical networks (SD-EONs) are more challenging than that in the conventional optical networks because failures can affect control plane operations. Meanwhile, traffic grooming is enabled by sliceable transponders can reduce power consumption and obtain higher spectrum efficiency. In this paper, we study the survivable grooming routing and spectrum allocation in SD-EON. First, provide an integer linear programming formulation to minimize both the required transponders and the maximum number of occupied frequency slots. Then, we develop a heuristic algorithm called shared backup path grooming protection (SBPGP) to get enough protection and less resources consumption. Extensive simulations are performed to study the power consumption of optical elements in data plane. Numerical results show that the proposed SBPGP scheme achieves better performances than the traditional shared backup path protection without grooming.

To improve the energy conversion ability and well utilize renewable resources, J. Rifkin first put forward the concept of “Internet of energy IoE)”. Although a peer-to-peer energy sharing mechanism is achieved through bi-directional energy transportation, the approach to solving cooperative energy transportation and storage still needs improving. Traditionally, the redundant energy will be wasted if it cannot be consumed by power load. In fact, the redundant energy can be stored to supply power loads in the future. For this end, we investigate cooperative energy transportation and storage for IoEs in terms of problem analysis, algorithm design, and platform development. After demonstrating the feasibility condition and proving the NP-hard of our problem, we derive the optimal solution by the reduction from a classic knapsack problem. We also design novel heuristics followed by different energy storage strategies. In addition, based on software-defined networking (SDN), a complementary platform is developed to make an effective decision for cooperative energy transportation and storage using heuristics above. Both simulation and experimental results demonstrate the effectiveness of our solutions.

In this letter, we report a diode-end-pumped continuous wave (CW) and passively Q-switched Ho,Pr:LLF laser at 2.95 μm. A maximum CW output power of 98 mW was achieved. The highest output power of 58 mW was radiated with a pulse duration of 818.8 ns and a repetition rate of 71.05 kHz. Even shorter pulse duration of 731.5 ns was obtained at a minor output coupler transmission.

We describe passively Q-switched Er:LuAG laser characteristics at 1.65 μm for the first time, with MoS2 and WS2 as the saturable absorber, respectively. The saturable absorbers were both prepared with the liquid phase exfoliation method. A shortest pulse width of 1.1 μ s was attained with the pulse repetition frequency of 41.6 kHz for the MoS2 saturable absorber, resulting in the highest peak power of 4.33 W. For the WS2 saturable absorber, a minimum pulse of 1.6 μs was gained with the pulse repetition frequency of 34.2 kHz and the calculated peak power of 4.13 W.

People's pursuit of higher quality communication has never ceased, which challenges the layout of 5G networks. Physical-layer network coding (PNC), as a key technology for 5G, supplies a powerful platform through leveraging the broadcast nature of wireless media. However, the symbol error rate (SER) of PNC is not well investigated, which would seriously influence user's quality of experience due to packet loss in wireless environment. In this paper, considering both phase and symbol misalignments, we perform analysis on SER of asynchronous PNC. By assuming part of information is known to the relay, and we derive the lower bound for SER. Afterward, through applying the concept of error vector and eliminating redundant terms, we derive the upper bound for SER. Both the lower and upper bounds are applicable to either multiuser detection-based network coding or belief propagation-based maximum a posteriori decoding. Finally, Monte Carlo simulation verifies our results and demonstrates that the bounds are relatively tight. The analytical results derived in this paper can facilitate future studies of practical and theoretical issues on PNC.

Given the possibility of ubiquitous 5G wireless access, location data bridge the gap between the physical world and digital online social networking services and also reflect user preferences and even interdependence among users. Owing to this interdependence, the advertiser can push corresponding products to users according to location recommendations, i.e., advertisement targeting. To achieve this mobile cloud computing service, after moving the computing capacity away from end devices to data centres (DCs), the converged infrastructure integrating optical metro and ubiquitous wireless access technologies is proposed in accordance with the 5G model. To minimize energy consumption, we propose a novel design framework of location-recommendation-aware virtual network embedding. This design framework determines the interdependency among user groups so that we can embed the virtual networks owned by user groups with a high interdependence into the same DC of the substrate optical-wireless hybrid infrastructure. Thus, the adviser can locally push the corresponding products to user groups at a single DC instead of consuming a large amount of energy to build inter-DC paths. The simulation results show that our design framework has greater energy efficiency and more profitable advertisement targeting compared with the benchmark, and the result of our heuristic is very close to the upper bound.

The aerospace-based communications can be managed more efficiently through the construction of an integrated space/air information network by the convergence of satellite (space) and unmanned aerial vehicle (air) networks. Such an integrated network would best fit the advent of delay- and disruption-tolerant networking, in which the data transmission can tolerate long delay and disruption under a store-carry-forward mechanism. Such a network, however, has some challenging research needs due to the network's high mobility of nodes and time-varying topology that may result in high error bit rate and long delay. In this paper, we propose a unified routing framework for this integrated network, where a Hybrid time-space Graph supporting Hierarchical Routing (HGHR) algorithm is achieved. More specifically, the HGHR performs on a hybrid time-space graph, including two subgraphs: a deterministic graph for the space network and a semi-deterministic one for the air network. This latter graph is based on a discrete time homogeneous semi-Markov prediction model. The hybrid time-space graph is then transformed into a state-space graph, based on which, a message forwarding rule under the store-carry-forward mechanism is adopted. Simulation results show that the proposed HGHR algorithm has good performance in terms of message delivery ratio, end-to-end delay, and power consumption.