Search

The ability of a novel biorealistic hand prosthesis for grasp force control reveals improved neural compatibility between the human-prosthetic interaction. The primary purpose here was to validate a virtual training platform for amputee subjects and evaluate the respective roles of visual and tactile information in fundamental force control tasks. We developed a digital twin of tendon-driven prosthetic hand in the MuJoCo environment. Biorealistic controllers emulated a pair of antagonistic muscles controlling the index finger of the virtual hand by surface electromyographic (sEMG) signals from amputees’ residual forearm muscles. Grasp force information was transmitted to amputees through evoked tactile sensation (ETS) feedback. Six forearm amputees participated in force tracking and holding tasks under different feedback conditions or using their intact hands. Test results showed that visual feedback played a predominant role than ETS feedback in force tracking and holding tasks. However, in the absence of visual feedback during the force holding task, ETS feedback significantly enhanced motor performance compared to feedforward control alone. Thus, ETS feedback still supplied reliable sensory information to facilitate amputee’s ability of stable grasp force control. The effects of tactile and visual feedback on force control were subject-specific when both types of feedback were provided simultaneously. Amputees were able to integrate visual and tactile information to the biorealistic controllers and achieve a good sensorimotor performance in grasp force regulation. The virtual platform may provide a training paradigm for amputees to adapt the biorealistic hand controller and ETS feedback optimally.

Sensory feedback provides critical interactive information for the effective use of hand prostheses. Non-invasive neural interfaces allow convenient access to the sensory system, but they communicate a limited amount of sensory information. This study examined a novel approach that leverages a direct and natural sensory afferent pathway, and enables an evoked tactile sensation (ETS) of multiple digits in the projected finger map (PFM) of participants with forearm amputation non-invasively. A bidirectional prosthetic interface was constructed by integrating the non-invasive ETS-based feedback system into a commercial prosthetic hand. The pressure information of five fingers was encoded linearly by the pulse width modulation range of the buzz sensation. We showed that simultaneous perception of multiple digits allowed participants with forearm amputation to identify object length and compliance by using information about contact patterns and force intensity. The ETS enhanced the grasp-and-transport performance of participants with and without prior experience of prosthetic use. The functional test of transport-and-identification further revealed improved execution in classifying object size and compliance using ETS-based feedback. Results demonstrated that the ETS is capable of communicating somatotopically compatible information to participants efficiently, and improves sensory discrimination and closed-loop prosthetic control. This non-invasive sensory interface may establish a viable way to restore sensory ability for prosthetic users who experience the phenomenon of PFM.

To address the urgent need to improve the ship type discrimination capability in maritime target detection, this article explores synthetic aperture radar (SAR) three-dimensional (3-D) imaging of moving ships, which can provide high robustness and stability of 3-D structural characteristics for ship type discrimination. This paper defines the three-degree-of-freedom (3-DOF) micro-Doppler frequency of ship scatterers, specifically referring to the Doppler frequency caused by the three-axis rotation of the ship when the motion of the radar platform and the ship's planar motion are both compensated. The 3-DOF micro-Doppler frequency of the ship scatterers is closely related to the roll, pitch, and yaw three-axis rotation parameters of the ship and contains the 3-D coordinate information of the scatterers. Building upon the high-precision extraction of the 3-DOF micro-Doppler frequency of ship scatterers, this paper proposes an SAR 3-D imaging method for moving ships based on this frequency. The proposed method describes the ship's rotational characteristics using 27 parameters and converts the variation in the 3-DOF micro-Doppler frequency of the scatterer into a set of weighted coefficients of B-spline basis functions. Two long short-term memory networks are constructed to estimate the 27 parameters describing the ship's rotation and invert the 3-D coordinates of the scatterers. Finally, through the processing of simulated and experimental data, the feasibility of using the 3-DOF micro-Doppler frequency of ship scatterers to achieve high-precision estimation of the ship's three-axis rotation parameters and high-precision inversion of 3-D coordinates was verified.

This study presents an innovative approach for estimating the azimuth cutoff wavelength ($\lambda _{c}$) using a multipolarization combination technique to enhance the retrieval of significant wave height (SWH) and wind speed (WS) from Gaofen-3 (GF-3) SAR wave mode data. The study identifies distinct advantages of copolarization for low to moderate sea states and cross-polarization for high sea states in the $\lambda _{c}$ estimation. Consequently, a suite of dual and quad-polarization combination methods is proposed, with the VV+VH combination demonstrating superior cost-efficiency, reducing the root mean square error (RMSE) of $\lambda _{c}$ estimation by around 20% compared with VV polarization. Correlation analysis between $\lambda _{c}$ at various polarizations, particularly VV+VH, and factors such as SWH, WS, wind direction, wave direction, and incidence angle, indicates a strong positive relationship with SWH and WS, and a moderate relationship with incidence angle. This insight informs the development of three $\lambda _{c}$-based SWH and WS retrieval models: single linear regression, multiple linear regression (MLR), and Gaussian process regression (GPR). The MLR and GPR models integrate normalized radar cross section (NRCS) and incidence angle to improve retrieval accuracy. The GPR model achieves more accurate estimation of SWH and WS compared with existing $\lambda _{c}$-based algorithms, with an RMSE of 0.485 m for SWH retrieval and 1.390 m/s for WS retrieval. Despite the performance gap with state-of-the-art algorithms, the GPR model offers exceptional cost-effectiveness and surpasses NRCS-based models for WS retrieval without requiring wind direction input.

Transient simulation in power engineering is crucial as it models the dynamic behavior of power systems during sudden events like faults or short circuits. Electromagnetic transient simulations involve multiple coordinated tasks. Traditional simulations are centralized and struggle to meet scalability requirements. To achieve these goals, distributed electromagnetic transient simulation has emerged as a new trend. Nevertheless, the distributed electromagnetic transient simulation introduces network communication. Achieving real-time simulation across distributed nodes poses the challenge of minimizing communication costs. In this paper, our proposal focuses on optimizing the task orchestration to reduce communication costs. Specifically, in the electromagnetic transient simulation, these tasks has certain communication pattern where the communicated objects of each task are pre-defined. We represent the pattern as a graph, with tasks represented as nodes and communications as edges. Furthermore, we propose to use graph partition with the objective of minimal communication costs and fine tune the partitions with the resource requirements of each distributed node. The experimental results demonstrate that our proposal has strength in achieving high-performance electromagnetic transient simulation.

Synthetic aperture radar (SAR) is widely used for observing sea surfaces and retrieving 2-D wave spectra. However, existing methods for retrieving directional wave spectra from SAR imagettes face challenges due to the complex nonlinear SAR-wave imaging relationship and the limitation of first-guess spectra. This study proposes a novel two-stage machine learning strategy for retrieving 2-D directional wave spectra from Chinese Gaofen-3 SAR wave mode products. We achieve the generation of complete 2-D wave spectra and several wave parameters solely from GF-3 SAR data without necessitating any additional inputs. In the first stage, we employ the Energy Attention Conditional Generative Adversarial Network (EA-CGAN) to retrieve the normalized wave spectrum. The generator of the EA-CGAN establishes a nonlinear transformation from normalized SAR cross spectra to normalized wave spectra to enhance the capabilities. In the second stage, the XGBoost model retrieves the intensity of the wave spectrum. The EA-CGAN and XGBoost models were trained on an extensive dataset that consists of about 11 000 Gaofen-3 SAR wave mode imagettes and 2-D wave spectra from the fifth-generation reanalysis (ERA-5) of the European Centre for Medium-Range Weather Forecasts. The results of the evaluation using test samples reveal high consistency between the retrieved wave spectra and ERA-5 wave spectra in terms of spectral similarity, peak period, peak direction, significant wave height, and mean wave periods. Compared to the traditional methods, our approach offers enhanced effectiveness, demonstrating the potential of advanced deep learning in high-precision SAR wave spectrum inversion.

The present study utilizes the backward propagation neural network (BPNN) and the dual-branch convolutional neural network (DB-CNN) algorithms to construct models for estimating significant wave height (SWH) from polarimetric Gaofen-3 SAR wave mode data, using a dataset of 11 164 images that are collocated with SWH from the European Centre for Medium-Range Weather Forecasts fifth generation reanalysis (ERA5). The models are assessed and compared across nine polarizations [vertical–vertical (VV), horizontal–horizontal (HH), RL, vertical–horizontal (VH), horizontal-vertical (HV), RR, 45° linear, RV, and RH] and various sea states using the SAR-ERA5 test samples as well as buoy and altimeter SWH observations. The results demonstrate the robust performance of BPNN models, with RMSEs around 0.30–0.32 m on SAR-ERA5 test data, 0.32–0.48 m on buoy data, 0.40–0.48 m on Jason-3 data, and 0.36–0.42 m on SARAL data. By comparison, the DB-CNN models, which additionally include two-dimensional (2-D) image spectra as input, only exhibit improved performance at VV, 45° linear, and RL polarizations, while showing negligible improvement at HH, RV, and RH polarizations and a notable degradation at VH, HV, and RR polarizations. Furthermore, the DB-CNN models generally fail to improve the overestimation (underestimation) in low (high) seas, and they even aggravate the overestimation (underestimation) under most polarizations. Additionally considering the heightened complexity, increased vulnerability to overfitting, and training times that are more than 200 times longer, the use of complex deep learning network structures to incorporate 2-D spectral information appears to be operationally limited for SAR SWH retrieval.

A convolutional neural network combined with long short-term memory (CNN-LSTM) phase compensation method (PCM) is proposed and demonstrated, where CNN is employed to extract spatial features, and LSTM is used to capture temporal features and realize the long-term predictions of residual phase fluctuations. This is the first-time machine learning (ML) has been used to mitigate the effects of optical path asymmetry caused by temperature variations on radio frequency (RF) transmission systems. The performance is verified by experiments on a unidirectional two-way RF transmission system, in which both the two 259-km-long separate fibers are coupled into one optical cable. The results demonstrate the CNN-LSTM model presents better prediction performance than the other eight previously proposed ML models. When the prediction duration is 40,000 s and the ambient temperature variation range is 14.38 °C, the coefficient of determination (R Squared, R2) between the predicted value and the actual value is higher than 0.99. In addition, compared to the phase locked loop (PLL) PCM, the proposed CNN-LSTM PCM can reduce the root-mean-square (RMS) phase jitter of the received signal from 219 ps to 19.72 ps, and improve the frequency stability of the system at 10,000 s by 84.5%. Overall, the proposed CNN-LSTM PCM can effectively compensate for residual phase fluctuations generated by the optical path asymmetry, providing a potential option for achieving stable RF transmission in telecommunication networks.

Computational biomechanical analysis plays a pivotal role in understanding and improving human movements and physical functions. Although physics-based modeling methods can interpret the dynamic interaction between the neural drive to muscle dynamics and joint kinematics, they suffer from high computational latency. In recent years, data-driven methods have emerged as a promising alternative due to their fast execution speed, but label information is still required during training, which is not easy to acquire in practice. To tackle these issues, this paper presents a novel physics-informed deep learning method to predict muscle forces without any label information during model training. In addition, the proposed method could also identify personalized muscle-tendon parameters. To achieve this, the Hill muscle model-based forward dynamics is embedded into the deep neural network as the additional loss to further regulate the behavior of the deep neural network. Experimental validations on the wrist joint from six healthy subjects are performed, and a fully connected neural network (FNN) is selected to implement the proposed method. The predicted results of muscle forces show comparable or even lower root mean square error (RMSE) and higher coefficient of determination compared with baseline methods, which have to use the labeled surface electromyography (sEMG) signals, and it can also identify muscle-tendon parameters accurately, demonstrating the effectiveness of the proposed physics-informed deep learning method.

Stabilizing the CMOS-compatible low-temperature Au-free GaN Ohmic contact is a critical work that determines the performance and yield of GaN power HEMTs in mass production. The instability of this contact has been puzzling the industry and academia for years. In this work, an overlooked factor, fluorine injection, is unambiguously verified to widely exist during dielectric etching and can easily destroy the low-temperature GaN Ohmic contact formation. The injection depth is confirmed to be over 30 nm with a fluorine peak concentration of 1023 at/cm3 in vicinity of the surface. Traditional method of partial AlGaN recessing with a pretty tiny processing window is proven unfriendly for production and vulnerable to the fluorine injection. Two methods to get rid of the fluorine are proposed. The first one is to over-etch the AlGaN barrier to the GaN channel to fully remove the fluorine ions. The second is to deposit an etch-stop blocking layer of Al2O3, which is also compatible with CMOS process.

This paper studies the distributed bipartite output formation control problem of heterogeneous discrete-time linear multi-agent systems (MASs) via cooperative output regulation theory. To construct the bipartite formation, under the structurally-balanced augmented directed signed graph, the followers of two antagonistic subgroups are supposed to respectively keep the desired relative positions with the leader, also called as the exosystem, in the same magnitude but the opposite sign. Since the information of exosystem can not be directly obtained by all followers, the distributed exosystem observer based on the discrete-time algebraic Riccati equality (ARE) is presented and the distributed state feedback controller is further designed. Moreover, in the case where the states of followers are not available, the distributed output feedback controller is proposed by introducing the state estimator. Finally, two numerical examples are given to demonstrate the effectiveness of analytic results.

Retinal disease detection and diagnosis relying solely on artificial retinal diseases will put great pressure on doctors and increase the rate of misdiagnosis. Therefore, the development of computer vision technology has brought the possibility for ophthalmologists to use computer-aided diagnosis. In recent years, most models for retinal disease recognition have been based on deep learning, which has the disadvantage of requiring large amounts of data and training time. It is also partly based on broad learning and its disadvantages are that feature extraction ability is limited and poor scalability. To overcome these limitations, we propose a novel artificial intelligence-based approach for the automatic assessment of retinal disease images called a multi-view deep-broad learning network (MDBL-Net), which absorbs the advantages of deep learning and broad learning. MDBL-Net comprises a Multi-view and Multi-scale Feature Extraction (MMFE) module and a Multi-scale Aggregation (MA) block. The MMFE module extracts features of different scales by learning feature representations from multiple views, while the MA block fully aggregates multi-scale deep-broad features from low-level to high-level representations. Experiments were conducted on two public datasets, UCSD and OCT2017. Experiments were conducted on two public datasets, UCSD and OCT2017, and results demonstrate that MDBL-Net achieves high accuracy even with limited training data (only 1%) and significantly reduces training time compared to traditional deep learning models. Specifically, MDBL-Net achieved an accuracy of 96.93% on the UCSD dataset and 99.90% on the OCT2017 dataset, outperforming state-of-the-art models in both cases. These findings suggest that the proposed MDBL-Net approach holds great promise for the task of retinal disease screening and recognition.

Multitemporal change detection (CD) plays a crucial role in the remote sensing application field. In recent years, supervised deep learning methods have shown excellent performance in detecting changes in very-high-resolution (VHR) images. However, these methods require a large number of labeled samples for training, making the process time-consuming and labor-intensive. Unsupervised approaches are more attractive in practical applications since they can produce a CD map without relying on any ground reference or prior knowledge. In this article, we propose a novel unsupervised CD approach, named transformer-based multivariate alteration detection (trans-MAD). It utilizes a pre-detection strategy that combines the compressed change vector analysis and the iteratively reweighted multivariate alteration detection (IR-MAD) to generate reliable pseudotraining samples. More accurate and robust CD results can be achieved by leveraging the IR-MAD to detect insignificant changes and by incorporating the transformer-based attention mechanism to model the difference or similarity between two distant pixels in an image. The proposed trans-MAD approach was validated on two VHR bitemporal satellite remote sensing datasets, and the obtained experimental results demonstrated its superiority comparing with the state-of-the-art unsupervised CD methods.

In the context of global warming, the lake surface water temperature (LSWT) exhibits a general upward trend. As an indicator of climate change, the surface water temperature of plateau lakes is particularly sensitive to climate warming. In this study, the machine learning model and physical process model are combined to predict water temperature in lake systems: The Air2water model and the long short term memory (LSTM) model are combined to form a prediction model for the inversion and prediction of 11 typical plateau lakes in the Yunnan–Guizhou Plateau. The modeling results show that the Air2water model performs best, followed by the LSTM model. Overall, the Air2water model and the LSTM model effectively reproduce the monthly, seasonal, and interannual variations of LSWT dynamics in the 11 lakes. The research results show that the LSWT dataset from 2021 to 2025 is constructed based on the forecast result, and it is found that six lakes, including Caohai, Dianchi, Erhai, Qionghai, Xingyun Lake, and Yangzonghai, showed significant warming in the next five years (the maximum warming rate was less than +0.2 °C/year), while the remaining five lakes showed no significant changes, and the size classification of LSWT data in different spaces on the interannual scale was consistent with the classification of altitudes. By conducting extensive exploration and research on the inversion of LSWT using physical process models and machine learning models, this study offers novel solutions for LSWT prediction and inversion. The study also provides a solid theoretical and practical foundation for future LSWT research, thereby holding practical significance and research value.

In recent years, deep learning has received tremendous attention in the field of hyperspectral unmixing (HU) due to its powerful learning capabilities. Particularly, the unsupervised unmixing method based on an autoencoder (AE) has become a research hotspot. Most of the current AE unmixing networks mainly focus on information about pixels and their neighborhoods in images. However, they make insufficient use of information about spatial heterogeneity and spectral differences of endmembers in hyperspectral image (HSI) data. To this end, an AE HU network with the name of SSF-Net is proposed for fusing the spatial–spectral features. The network first extracts pseudoendmember information from the HSI using a regional vertex component analysis algorithm. Then, a dual-branch feature fusion module incorporating a spatial–spectral attention mechanism is constructed to make full use of the information in the HSI data, thereby improving the network's unmixing performance. It is worth stating that SSF-Net can fuse spatial–spectral information and utilize different attention maps to obtain more significant spectral difference information and more discriminative spatial difference information about the scene. The experimental results on synthetic and real datasets demonstrate that the proposed SSF-Net outperforms state-of-the-art unmixing algorithms.

The global context is crucial to the semantic segmentation task of remote sensing (RS) urban scene imagery since objects have large size variations, high similarity, and mutual occlusion. However, the existing methods for extracting global context information have limitations when directly applied to very high-resolution RS images, mainly in high complexity of computation and memory consumption. To alleviate this limitation, we propose a novel Efficient Global Understanding semantic segmentation Network (EGUNet) to extract global context information efficiently for applicability to RS images. Specifically, EGUNet is a hybrid U-shaped architecture of convolutional neural networks (CNNs) and Transformer in which the encoder uses the CSWin Transformer to capture global semantic information, and the decoder uses the CNNs structure to recover local detail information. Thus, the proposed EGUNet has a powerful global extraction capability and local position information recovery capability. In addition, three effective modules are proposed to improve the segmentation accuracy to make EGUNet more applicable for urban scene image segmentation tasks. First, a feature adaptive fusion module is introduced in the decoder to improve the fusion of the deep semantics and the location detail features. Second, an adaptive atrous-spatial pyramid pooling is designed at the skip connections to enhance the multiscale understanding of high-level semantic context. Finally, we introduce a lightweight enhanced segmentation head to utilize the information from each decoder stage for segmentation. Extensive experimental results on ISPRS Vaihingen and Potsdam datasets demonstrate the exceptional segmentation accuracy of EGUNet, outperforming the state-of-the-art methods.

The Tianwen-1 mission, China's first interplanetary endeavor and Mars mission, touched the surface of the Red Planet on May 15, 2021. With the successful landing of the Zhurong rover on the southern Utopian Planitia of Mars, the Mars Surface Composition Detector (MarSCoDe) on board the rover has started to analyze the material composition of the Martian surface by using the laser-induced breakdown spectroscopy (LIBS). However, changes in instrument temperature and external environment during operation will cause spectral drift in LIBS data, leading to inaccurate material composition inversion results due to shifts in elemental characteristic peak positions. To address this problem, an adaptive spectral drift correction (ASDC) approach is proposed. By considering the distribution of characteristic peaks and drift differences between different elements, the proposed ASDC method corrects the spectral drift in LIBS science data according to an adaptive spectral segmentation strategy. Experimental results obtained on 36 real LIBS science data acquired over 19 exploration activities confirmed the effectiveness of the proposed approach in terms of lower average drift values, by comparing it with two reference methods. Taking the onboard calibration data as a reference, the average spectral drift values of the LIBS science data before and after applying ASDC decreased from 0.1012 nm to 0.0179 nm, 0.0926 nm to 0.0158 nm, and 0.1697 nm to 0.0103 nm in three LIBS channels, respectively. Furthermore, the resulting decrease in the standard deviation and rms error of the average drift values also proved the robustness of the proposed method.

Owing to the strong protection of data privacy, federated learning (FL) has become a key method to achieve intelligent decision making in smart homes. However, under realistic conditions, such as differentiated requirements and heterogeneous service environments, FL in smart homes faces the problems of non-independent and identically distributed (non-IID) data and uneven computing power, which lead to the poor adaptability of global models. To address this issue, this study proposes a cluster FL architecture based on edge-cloud collaboration. Firstly, a Gaussian mixture model-based cluster FL is proposed to improve the model accuracy by clustering features on the FL dataset and ensuring an independent identical distribution of the data. Subsequently, a model training strategy based on edge-cloud collaboration is proposed to achieve the sharing of edge-cloud computing power by split learning, which provides sufficient computing power for model training. The simulation results show that the proposed architecture improves the accuracy of global models, meanwhile ensures the provision of normal network service.

Substation equipment defects are important factors affecting the safe operation of power grids. However, many non-rigid defects have low detection accuracy and poor robustness,due to boundary ambiguity, irregular shape and tiny size. To address these problems,we propose a swin-transformer with recursive gated convolution framework for substation equipment non-rigid defect. Firstly, in order to effectively detect non-rigid defect objects to improve the discriminability of image features, we design the Swin-Transformer with Recursive Gated Convolution(Swin-RGC) framework to extract the interaction features between spaces in the deep model. Secondly, to avoid the loss of object location information, the Task-aligned One-stage Object Detection(TOOD) head is improved by fusing Coordinate Attention modules. Finally, a substation equipment defect detection dataset is established to provide a baseline for detecting non-rigid defects in substation power equipment. Experiment results on our dataset demonstrate that our proposed method achieves the performance of 69.9% Mean Average Precision (mAP) in the substation equipment non-rigid defect detection, which outweighs the state-of-the-art approaches.

Musculoskeletal models have been widely used for detailed biomechanical analysis to characterise various functional impairments given their ability to estimate movement variables (i.e., muscle forces and joint moments) which cannot be readily measured in vivo. Physics-based computational neuromusculoskeletal models can interpret the dynamic interaction between neural drive to muscles, muscle dynamics, body and joint kinematics and kinetics. Still, such set of solutions suffers from slowness, especially for the complex models, hindering the utility in real-time applications. In recent years, data-driven methods have emerged as a promising alternative due to the benefits in speedy and simple implementation, but they cannot reflect the underlying neuromechanical processes. This paper proposes a physics-informed deep learning framework for musculoskeletal modelling, where physics-based domain knowledge is brought into the data-driven model as soft constraints to penalise/regularise the data-driven model. We use the synchronous muscle forces and joint kinematics prediction from surface electromyogram (sEMG) as the exemplar to illustrate the proposed framework. Convolutional neural network (CNN) is employed as the deep neural network to implement the proposed framework. Simultaneously, the physics law between muscle forces and joint kinematics is used the soft constraint. Experimental validations on two groups of data, including one benchmark dataset and one self-collected dataset from six healthy subjects, are performed. The experimental results demonstrate the effectiveness and robustness of the proposed framework.

This paper addresses the distributed secure consensus problem for linear multi-agent systems (MAS) with unmeasurable states and Denial-of-Service (DoS) attacks. The observer is designed to be used in estimating unmeasurable states of agents. Then, an observer-based distributed event-triggered control protocol is proposed to save network resources. By using the piecewise Lyapunov function approach, several sufficient conditions for ensuring secure consensus are obtained. Further, a rigorous theoretical analysis shows that the given event-triggering scheme exhibits no Zeno behavior. Finally, the effectiveness of the proposed method is verified by numerical simulation.

Studying the changes of vegetation coverage and carbon storage in a terrestrial ecosystem will provide important information for ecosystem optimization. The Kubuqi Desert is a typical ecological fragile area, and its ecological construction is not only to build an ecological barrier to protect northern China but also to be a case of similar regions in the world. Engebei ecological demonstration area, located on the northern margin of the central Kubuqi Desert, was selected as the study area. The objective is to establish a carbon inversion model through the LASSO regression model in this area which can promote its ecological development. Taking Landsat, Sentinel-2A remote sensing data, and field sample data as data sources. The vegetation coverage was calculated through the remote sensing data. The carbon reserves in plots were calculated through the biomass equation. Then, the appropriate carbon storage estimation model was found, which was established by the LASSO regression model and the Sentinel-2A remote sensing images. Finally, the carbon reserves over the years were calculated by this estimation model. Research shows that: first, the overall vegetation coverage of the study area was not high, which was around 27%–37%; second, the proportion of lower (I) level vegetation coverage in the study area decreased overall, and the proportion of higher (V) level vegetation coverage generally stabilized; third, carbon reserves remain roughly above 71752.64t, up by 20,209.61t in 2021 compared to 2020, and up by 14,867.02t in 2021 compared to 1991.

Hyperspectral image (HSI) classification has always drawn great attention in the field of remote sensing. Various deep learning models are in the ascendant and gradually applied to HSI classification. Nevertheless, limited-labeled and class-imbalanced datasets largely make the classifier prone to overfitting. To address the above problem, this article proposes a two-branch generative adversarial network with multiscale connections (TBGAN), which includes two generators to produce the spectral and spatial samples, respectively. Thereinto, the spectral generator is imbued with the self-attention mechanism to maximumly capture the long-term dependencies across the spectral bands. And meanwhile, an elaborated discriminator with two branches is devised in TBGAN for extracting the joint spectral-spatial features. Besides, the multiscale connections are placed between the discriminator and two generators to alleviate the instability problems caused by the inherently backward propagation of gradients in GAN. Furthermore, a feature-matching term is added to the loss function to prevent the generators from overtraining upon the current discriminator, thereby further improving the stability of the network. Experiments upon three benchmark datasets demonstrate that TBGAN achieves an extremely competitive classification accuracy and exerts lower sensitivity to the training sample size compared with several state-of-the-art methods.

By integrating communications in different domains, integrated radio and optical networks can serve a wider range of applications and services. Integrated radio and optical network scenarios will involve more weak-computation-ability network nodes, such as small-cell base stations. To pursue efficient integrated radio and optical networks, more efficient ways to conduct transmission under the demand of edge and cloud collaboration are required. The lack of forward-looking resource allocation may easily lead to a waste of network resources without an expected return. Therefore, an efficient resource allocation scheme needs to consider certain issues: 1) a comprehensive perspective of traffic prediction; 2) a release of pressure on the transmission pipeline during the prediction process; and 3) a reduction of loss of edge nodes due to the computation. In this paper, benefiting from machine learning, we propose a resource allocation with edge-cloud collaborative traffic prediction (TP-ECC) in integrated radio and optical networks, where an efficient resource allocation scheme (ERAS) is designed based on the prediction results with the gated recurrent unit model. We maximize the utilization of limited resources to improve the awareness of network status. We present three evaluation indicators and build a network architecture to evaluate our resource allocation scheme. Through edge-cloud collaboration, our proposal can improve traffic prediction accuracy by 9.5% compared with single-point traffic prediction, and resource utilization is also improved by edge-cloud collaborative traffic prediction.

The attention mechanism of the Transformer has the advantage of extracting feature correlation in the long-sequence data and visualizing the model. As time-series data, the spatial and temporal dependencies of the EEG signals between the time points and the different channels contain important information for accurate classification. So far, Transformer-based approaches have not been widely explored in motor-imagery EEG classification and visualization, especially lacking general models based on cross-individual validation. Taking advantage of the Transformer model and the spatial-temporal characteristics of the EEG signals, we designed Transformer-based models for classifications of motor imagery EEG based on the PhysioNet dataset. With 3s EEG data, our models obtained the best classification accuracy of 83.31%, 74.44%, and 64.22% on two-, three-, and four-class motor-imagery tasks in cross-individual validation, which outperformed other state-of-the-art models by 0.88%, 2.11%, and 1.06%. The inclusion of the positional embedding modules in the Transformer could improve the EEG classification performance. Furthermore, the visualization results of attention weights provided insights into the working mechanism of the Transformer-based networks during motor imagery tasks. The topography of the attention weights revealed a pattern of event-related desynchronization (ERD) which was consistent with the results from the spectral analysis of Mu and beta rhythm over the sensorimotor areas. Together, our deep learning methods not only provide novel and powerful tools for classifying and understanding EEG data but also have broad applications for brain-computer interface (BCI) systems.

Net load uncertainty in electricity spot markets is rapidly growing. There are five general approaches by which system operators and market participants can use probabilistic forecasts of wind, solar, and load to help manage this uncertainty. These include operator situation awareness, resource risk hedging, reserves procurement, definition of contingencies, and explicit stochastic optimization. We review these approaches, and then provide a case study in which a method for using probabilistic solar forecasts to define needs for reserves is developed and evaluated. The case study has three parts. First, we describe building blocks for enhancing the Watt-Sun solar forecasting system to produce probabilistic irradiance and power forecasts. Second, relationships between Watt-Sun forecasts for multiple sites in California and the system’s need for flexible ramp capability (flexiramp) are defined by machine learning and statistical methods. Third, the performance of present methods to defining flexiramp requirements, which are not conditioned on weather and renewables forecasts, is compared with that of probabilistic solar forecast-based requirements, using a multi-timescale production costing model with an 1820-bus representation of the WECC power system. Significant potential savings in fuel and flexiramp procurement costs from using solar-informed reserve requirements are found.

Objective: Human neuromuscular reflex control provides a biological model for a compliant hand prosthesis. Here we present a computational approach to understanding the emerging human-like compliance, force and position control, and stiffness adaptation in a prosthetic hand with a replica of human neuromuscular reflex. Methods: A virtual twin of prosthetic hand was constructed in the MuJoCo environment with a tendon-driven anthropomorphic hand structure. Biorealistic mathematic models of muscle, spindle, spiking-neurons and monosynaptic reflex were implemented in neuromorphic chips to drive the virtual hand for real-time control. Results: Simulation showed that the virtual hand acquired human-like ability to control fingertip position, force and stiffness for grasp, as well as the capacity to interact with soft objects by adaptively adjusting hand stiffness. Conclusion: The biorealistic neuromorphic reflex model restores human-like neuromuscular properties for hand prosthesis to interact with soft objects.

Human activity recognition (HAR) is gaining interest with many important applications including ubiquitous computing, health-care services and detection of diseases. Smartphone sensors have high acceptance and adherence in daily life and they provide an alternative and economic way for activity recognition. To improve the performance of smartphone sensor-based HAR, a novel smartphone sensor-based HAR method (hybrid diversity enhancement with selective ensemble learning, HDESEN) that utilizes selective ensemble learning with differentiated extreme learning machines (ELMs) is proposed, where hybrid diversity enhancement is proposed to boost the diversity of base models and an improved binary glowworm swarm optimization (IBGSO) is employed to effectively enhance the learning process by choosing a superior subset for ensemble instead of all. Firstly, statistical features in the time domain and frequency domain are extracted and integrated from smartphone sensors and then three filter-based feature selection methods are utilized for desirable base models. Secondly, to enhance the diversity of the base models, three types of diversities are introduced to construct different base models, respectively. Among them, Bootstrap is introduced to design distinctive training data subsets for differential base models, random subspace and optimized subspace are proposed to obtain different feature spaces for constructing base models. Thirdly, a pruning method based on glowworm swarm optimization (GSO) is proposed to find the optimal sub-ensemble from the pool of models from all diverse types to implement selective ensemble learning. The experimental results on tow publicly available datasets (UCI-HAR and WISDM) demonstrate the proposed HDESEN can reliably improve the performance of HAR and outperforms the relevant state-of-the-art approaches.

The operation characteristics of energy storage can help the distribution network absorb more renewable energy while improving the safety and economy of the power system. Mobile energy storage systems (MESSs) have a broad application market compared with stationary energy storage systems and electric vehicles due to their flexible mobility and good dispatch ability. However, when urban traffic flows rise, the congested traffic environment will prolong the transit time of MESS, which will ultimately affect the operation state of the power networks and the economic benefits of MESS. This paper proposes a bi-level optimization model for the economic operation of MESS in coupled transportation-power networks, considering road congestion and the operation constraints of the power networks. The upper-level model depicts the daily operation scheme of MESS devised by the distribution network operator (DNO) in order to maximize the total revenue of the system. With fuzzy time windows and fuzzy road congestion indexes, the lower-level model optimizes the route for the transit problem of MESS. Therefore, road congestion that affects the transit time of MESS can be fully incorporated in the optimal operation scheme. Both the IEEE 33-bus distribution network and the 29-node transportation network are used to verify and examine the effectiveness of the proposed model. The simulation results demonstrate that the operation scheme of MESS will avoid the congestion period when considering road congestion. Besides, the transit energy consumption and the impact of the traffic environment on the economic benefits of MESS can be reduced.

The uncertainty of user-side resource response will affect the response quality and economic benefit of load aggregator (LA). Therefore, this paper regards the flexible user-side resources as a virtual energy storage (VES), and uses the traditional narrow sense energy storage (NSES) to alleviate the uncertainty of VES. In order to further enhance the competitive advantage of LA in electricity market transactions, the opertion mechanism of LA in day-ahead and real-time market is analyzed, respectively. Besides, truncated normal distribution is used to simulate the response accuracy of VES, and the response model of NSES is constructed at the same time. Then, the hierarchical market access index (HMAI) is introduced to quantify the risk of LA being eliminated in the market competition. Finally, combined with the priority response strategy of VES and HMAI, the capacity allocation model of NSES is established. As the capacity model is nonlinear, Monte Carlo simulation and adaptive particle swarm optimization algorithm are used to solve it. In order to verify the effectiveness of the model, the data from PJM market in the United States is used for testing. Simulation results show that the model established can provide the effective NSES capacity allocation strategy for LA to compensate the uncertainty of VES response, and the economic benefit of LA can be increased by 52.2% at its maximum. Through the reasonable NSES capacity allocation, LA is encouraged to improve its own resource level, thus forming a virtuous circle of market competition.

As artificial intelligence continues to advance, deep learning has greatly contributed to the advancement of ship recognition using synthetic aperture radar (SAR) images. Deep learning-based SAR ship recognition performance is largely dependent on the sample set used. SAR ship recognition datasets published in recent years, however, are most derived from a single SAR satellite sensor. It needs to be evaluated and analyzed carefully whether the model trained by a single satellite dataset can still achieve the same accuracy with different SAR satellites. This article focuses on the following research to address these issues. First, using multiple SAR satellite sensors, we create a new SAR ship dataset (named generalization performance evaluation dataset, GPED) containing multiresolution and multipolarization data to examine the generalization performance of the deep learning-based SAR ship recognition method. GPED and a marine target detection dataset (MTDD) are then used to evaluate and analyze the generalization performance of current mainstream deep learning methods. According to the the experiment results, the mean average accuracy of the ship recognition model trained on GPED is generally higher than that of MTDD, which proves that the GPED has a better generalization performance. Furthermore, SAR ship detection datasets have more samples than ship recognition datasets, which inspired us to use transfer learning to transfer knowledge from ship detection to ship recognition. In this article, a method for ship recognition based on transfer learning that utilizes the knowledge gained from the ship detection task is proposed. The method includes two modules: 1) pretraining module and 2) fine-tuning module. It can apply samples of unlabeled ship types to ship recognition, thus reducing the number of labeled samples that are required for ship recognition. The experimental results on GPED and MTDD show that our method can achieve good recognition performances.

The rolling, pitching, and yawing motion of the vessel produces a micro-Doppler signature in the radar echoes. Ship scatterers’ three-degree-of-freedom (3-DOF) micromotion features contain information on the ship geometric structure and motion parameters. Therefore, it is essential to extract the 3-DOF micromotion features of ship scatterers with high accuracy. Because existing approaches to extracting 3-DOF micromotion features of ship scatterers under long accumulation time have the disadvantage of low accuracy, we propose a high-precision algorithm based on linear micro-Doppler trajectory tracking after performing a generalized S-transform and ridge extraction to coarsely analyze the time-frequency trajectory corresponding to the ship's echo. To improve the extraction accuracy of 3-DOF micromotion features of ship scatterers, long time is divided into many short time segments. The motivation for time segmentation is that the instantaneous Doppler frequency of a scatterer caused by a ship's rotation is verified to behave approximately as a linear frequency modulation signal over a short period. The Doppler centroids and frequency modulation rates of scatterers during a short time segment in a given range bin can be easily coarsely estimated from ridge extraction results. The instantaneous Doppler frequency trajectories of the same scatterer in adjacent short time segments are correlated and filtered by using the nearest neighbor algorithm and the extended Kalman filtering algorithm to achieve high estimation accuracy. The results of the performed simulation and processing of real radar data verify the accuracy of the proposed algorithm is higher than those of the current commonly used algorithms.

Outlier detection in vibration signals can play an important role in addressing the issue of structural or environmental changes during vibration testing. In this study, a transformer-based model for outlier detection is proposed. Unlike previous statistical and regression outlier detection methods, the proposed model can identify the outlier location in a high dimensional observation space using the self-attention mechanism. The location of outliers within the vibration observation is marked by a combination of a spatial label and a temporal label. The outlier detection performance of the model is verified by a numerical study of the plane wave and an experimental study of the vibrating plate. These two studies show that the proposed model has good label prediction accuracies (all above 85%) toward the outlier location within the plane wave and vibrating plate observations.

Founded on bounded rationality and limited information, evolutionary game theory has been preliminarily applied in many fields, such as electricity market (EM). To address the complex behavioral decision-making issues in the more-common three-population multi-strategy evolutionary game (3PmSEG) scenarios in EM. This paper explores the long-term evolutionarily stable equilibrium (ESE) characteristics of general 3PmSEG systems with the aim of systematically investigating the evolution process of long-term on-grid bidding of a generation-side EM based on these features. First, the long-term ESE characteristics of general three-population two-strategy and three-strategy evolutionary games are thoroughly investigated. Complete relative net payoff (RNP) parameters are defined for these games. Then, the modeling idea of general 3PmSEGs is elaborated. Research shows that the game can be guided to evolve toward an expected long-term ESE point by properly adjusting its RNP parameters. To verify this, finally, the long-term on-grid bidding of power generation is investigated for a tripartite generation-side EM. The case study reveals that effective government supervision can effectively promote new energy accommodation of the market. Overall, the models developed in this paper are relatively universal and practical, which can provide some theoretical and methodological references for complex evolutionary game issues in related fields.

Aiming to solving the nonlinear coordinated trajectory tracking control problem of multi-USVs under the influence of slowly changing and uncertain environmental disturbances, reducing energy consumption and ensuring tracking performance and stability, a distributed event-triggered adaptive coordinated trajectory tracking controller (DET-ACTTC) for multiple unmanned surface vessels (multi-USVs) is proposed based on the undirected communication topology. By introducing a virtual leader, a leader-follower formation is used to generate the coordinated tracking errors. Based on the expected formation and states, the aggregate tracking errors are defined, and event-triggered measurement errors are constructed to design the event-triggered conditions. An adaptive term is used to compensate for external environment disturbances, and the DET-ACTTC is deduced for each USV in the group. The tracking errors of the event-triggered coordinated trajectory tracking control (ETCTTC) system gradually converge to zero, and Zeno behavior is avoided. The theoretical results are verified by simulations.

In order to sail safely in a marine environment with neighboring obstacles, a local collision avoidance algorithm based on a steering maneuver considering International Regulations for Preventing Collisions at Sea (COLREGs) is proposed for unmanned surface vehicles (USVs). In this paper, the algorithm consists of three parts: collision risk assessment, steering occasion determination, and navigation waypoint update. The collision risk assessment is used to judge a collision risk based on an analytical feasible angle interval by the closest point of approach method. The steering occasion determination provides a safe distance interval from an obstacle by analyzing a marginal steering angular velocity. The navigation waypoint update generates a temporary waypoint to navigate the USV and considers the overtaking, head-on, and crossing rules of COLREGs. The temporary waypoint is calculated based on a real-time optimal orientation angle, which is determined by an optimization objective function for angle deviation. The three parts for local collision avoidance are determined by a finite state machine. In simulations, a USV with actual dynamic constraints runs successfully in a static and dynamic multiobstacle environment, and the results indicate the feasibility and validity of the proposed algorithm.

The reinforcement learning framework for multi-hop relational paths is one of the effective methods for solving knowledge graph incompletion. However, these models are associated with limited performances attributed to delayed rewards and far-fetched search trajectories. To overcome these challenges, we propose the searching window and multi-policy agent. The searching window provides a large action space, so that the agent can backtrack based on the newly obtained information and establish a local graph instead of a path chain. Based on the searching window, a double long short-term memory (DBL-LSTM) policy network is introduced to encode the local graph and relation sequence, after which the encoding information is used by the agent to select a correct entity to grow the local graph. Furthermore, multi-policy agent separately infers the local graph through three different policy networks, then, all local graphs are integrated into an information-rich local graph. Experiments using the WN18RR dataset revealed that local graph reasoning with searching window had greater rewards than path reasoning, the proposed DBL-LSTM policy network improved all HITS@N(N = 1,3,5,10) compared to prior works, and that the multi-policy agent achieved higher hit rates than single-policy agent.

The interferometric imaging radar altimeter (InIRA) is a new generation radar altimeter, which can provide two-dimensional images of the sea surface topography at high resolution along a wide swath. This article proposes a method for the simulation of sea surface images measured by InIRA. First, the Pierson-Moskowitz wave spectrum and two-scale model are used to simulate the sea surface from which elevation data is to be acquired. This simulated sea surface is then divided into small triangular facets using Delaunay triangulation. Second, the backscattering cross sections of these small facets are calculated via application of quasi-mirror scattering theory, and the backscattering coefficient of the simulated region derived via coherent superposition. Third, system parameters are set, consistent with the basic principle of InIRA. Assuming that the signal transmitted is a linear frequency modulation pulse signal, the simulation images are then derived using the range Doppler and back projection algorithms. By inverting the interferometric phase diagram, elevation estimates can be derived, and compared with original simulated sea levels. This demonstrated accuracy within the centimeter range, verifying the correctness and feasibility of the proposed method.

Accurate propagation characteristics are essential for future indoor millimeter-wave (mmWave) small cell network planning. This paper presents propagation measurements at 26 GHz and 38 GHz which are important candidate bands for fifth generation mmWave communication. Measurements are conducted in an indoor corridor, as well as a stairwell whose mmWave channel is seldom investigated before. In these measurements, an omnidirectional biconical antenna is used as transmitter and a steerable directional horn antenna is used as receiver. The directional and omnidirectional path loss exponents, shadow factors, cross-polarization discrimination ratios and root-mean-square delay spreads are analyzed for both line-of-sight and non-line-of-sight scenarios in both co-polarization and cross-polarization, and these characteristics are compared for different frequencies and environments. It is found obvious depolarization phenomenon in non-line-of-sight scenario for higher frequency. Compared to the corridor, the stairwell has larger path loss exponents and root-mean-square delay spreads, and the depolarization is also more evident in stairwell. The results in this paper are beneficial to building efficient and robust indoor mmWave communication systems.

Quantum-key-distribution (QKD) enabled optical networks can provide secure keys for different kinds of applications to ensure the high security of communication processes. As the number of wavelength channels in a single-core fiber is limited, space division multiplexing (SDM) has been considered to be a valuable technique to provide more wavelength resources, which solves the problem of insufficient channel capacity. Due to the rapid expansion of data volume, the SDM-based QKD network has been regarded as an important paradigm. At the same time, resource allocation in multi-core optical networks is a problem worthy of studying. In this work, four routing, core, and wavelength allocation (RCWA) algorithms are proposed in the multi-core optical network for quantum service requests and classical service requests, which are referred to as RCWA without perception (RCWA-WP), RCWA with crosstalk perception (RCWA-XTP), RCWA with spectrum perception (RCWA-SP), and RCWA with core perception (RCWA-CP). Simulations are carried out to evaluate the network performance in terms of blocking probability, key utilization, and average crosstalk intensity. The comparative results indicate that the RCWA-SP algorithm can improve network performance and reduce the blocking probability, while the RCWA-XTP algorithm can reduce inter-core crosstalk during transmission.

The human microbiome is comprised of thousands of microbial species. These species will substantially influence the normal physiology of humans and cause numerous diseases. Microbiome data can be measured by sequencing, microarray, or other technologies. With the fast development of these technologies, downstream analysis methods should also be designed to effectively and accurately discover the valuable information that is hidden in the data. Many methods have been designed for the count data of microbiome. However, to our knowledge, there are only a few methods developed for the continuous microbiome data. Many microbiome data have an over-dispersed and zero-inflated data structure. Traditional methods rarely characterize this data structure and only focus on the differences in the abundance between different samples. In this study, we introduce a novel method, the zero-inflated gamma (ZIG) omnibus test, to specifically test the continuous and zero-inflated microbiome data. In this test, abundance will be tested along with zero prevalence and dispersion. We compared this method with five other popular methods. We found that ZIG omnibus test has significantly higher power and a similar or lower false-positive rate than the competing methods in the tests of simulated data. It also found more proved microbiomes in the real data with tonsil cancer. So, we conclude that ZIG omnibus test is a robust method across various biological conditions in the differential expression test of microbiome data.

In recent years, sensor-based human activity recognition (HAR) has become a hot topic due to the advancement of sensing technologies, wireless communication technologies and nano-technologies. Since the sensor signals are usually non-stationary and quite noisy, both selecting the discriminant feature representations and finding out the optimal parameters for recognition algorithm play an important role for the enhanced performance and robustness of an HAR system. However, most of the previous research focused on one of them ignoring their interactions. Very few studies focused on these two aspects simultaneously. Considering the two factors separately may lead to inferior HAR performance. This paper presents a novel HAR framework which can optimize the feature set and the parameters of recognition algorithm synchronously for robust and optimal system performance. A new hybrid feature selection methodology using game-theory based feature selection (GTFS) and binary firefly algorithm (BFA), called GTFS-BFA, is proposed. GTFS-BFA is a hybrid methodology combining evidence from both filter and wrapper feature selection methods. It consists of two phases, namely pre-selection phase and re-selection phase. Pre-selection phase relies on game-theory-based filter method, while the re-selection phase uses binary firefly algorithm (BFA) as a wrapper method. The popular and efficient algorithm kernel extreme learning machine (KELM) is utilized as a classifier. The experimental results indicate that the proposed method can obtain better comprehensive performance in terms of four performance measures through a comparison to other existing methods on daily activity dataset from five body positions.

Graph convolutional neural networks (GCNs) have been widely used in hyperspectral images (HSIs) classification for their superiority in processing non-Euclidean structure data. The performance of GCNs relies on the initial graph structure. Most GCN models only utilize spectral information to construct a graph, which is inaccurate because they fail to take the relationship between adjacent nodes into consideration. In addition, due to the over-smooth phenomenon, most GCN models are shallow and unable to extract effective features. To address these issues, a dual graph u-nets is proposed by integrating spatial graph and spectral graph for HSIs classification, denoted by DGU-HSI. To integration the spectral and spatial information, two graphs are constructed for feature extraction. The spectral graph is created by spectral similarity among all pixels where multirange spectral information is retained, and the spatial graph is constructed by exploiting the neighborhood relationship of the center pixel, which describes spatial information. Then, a dual GCN is utilized to extract spatial and spectral graph features simultaneously. To relieve the over-smooth phenomenon, the graph u-nets structure is applied on constructed spectral and spatial graph to extract effective features. Finally, the extracted spectral and spatial features are fused for classification. Experiments conducted on the public datasets demonstrate the effectiveness of the proposed method on HSIs classification.

Accurate identification of sea ice type from altimeter data is important for inversion of sea ice freeboard by radar altimeter, as well as for maritime navigation in polar regions. Previous sea ice classification studies have rarely involved multifrequency altimeter data. In this article, benefiting from the dual-frequency observation capability of the Chinese HY-2B altimeter, differences between waveform features of C- and Ku-band radar altimeter over different types of sea ice were investigated and analyzed for the first time. To validate and evaluate the sea ice classification capability of different feature combinations of the different bands, we extracted 14 features from the waveforms of C- and Ku-band HY-2B altimeter data acquired between December 2019 and April 2020. Our experiments demonstrated that the C- and Ku-band radar waveforms are sensitive to sea ice type, and that sea ice type classification could be improved by combining Ku- and C-band data. This work also proposed a sea ice classification methodology using C- and Ku-band altimeter waveforms. Comparison with results from previous studies indicated that the overall accuracy of the proposed method (up to 93.45%) is much higher than that of existing state-of-the-art methods.

Time-Wavelength Division Multiplexing Passive Optical Network (TWDM-PON) is considered as a promising solution of next generation PON (NG-PON). The integration of Edge Computing (EC) and TWDM-PON can satisfy the QoS requirements of delay-sensitive applications by providing storage, processing and caching capabilities at the network edge. However, with limited resource capacity, edge nodes in TWDM-PON are vulnerable to network attacks, e.g., Distributed Denial of Service (DDoS) attacks. Resource exhaustion in the attacked nodes easily leads to QoS degradation and even service blocking. This paper investigates how to effectively schedule traffic to mitigate the DDoS attack in EC-enabled TWDM-PON. Based on the collaboration of edge nodes, an adaptive traffic scheduling algorithm is designed to minimize the impact of DDoS attacks on delay sensitive services. The performance of the proposed algorithm is evaluated in simulation, where direct and indirect DDoS attacks are simulated. Besides, the attack duration and the number of attacked nodes are considered in the evaluation. Simulation results show that the proposed algorithm can effectively mitigate DDoS attacks in terms of reducing the QoS degradation rate and blocking rate of delay-sensitive services.

A critical challenge in image restoration is the presence of various types of noise. Meanwhile, noise detection is a crucial step in mixed noise removal. This paper tackles the challenge of restoring images corrupted by a mixture of additive Gaussian and multiplicative Gamma noise. In the proposed method, we integrate the noise detection process into a variational model using a dual formulation of a maximum a posteriori (MAP) estimator. The variational model consists of a novel adaptive fidelity term and a plugin-and-play regularization term. The fidelity term contains an adaptive weight that can automatically detect the noise types, levels, and pollution ways for each pixel. There is flexibility in choosing a plugin-and-play regularization term. For example, we can use a model-based regularizer or a deep learning-based regularizer. In addition, we present a splitting algorithm to minimize the proposed cost functional. This splitting technique enables us to transfer a mixed noise removing problem to several subproblems, including noise removal and detection. The noise detection process can be iteratively estimated by the proposed algorithm itself. Therefore, in the numerical experiments, the proposed model outperforms the existing Rudin-Osher-Fatemi (ROF), Aubert-Aujol (AA), BM3D, and deep learning-based single type denoiser. Experimental results show that the proposed model can remove noise more efficiently and better preserve details in images. Compared to the existing best-performing single type denoiser, on average, the improvements of PSNR values range from 0.33 dB to 0.81 dB under noise mixture ratios $\alpha =0.4, 0.6$ .

Small cells (SCs) are expected to be ultra-densely deployed in or close to the traffic hot-spots in the fifth generation (5G) mobile networks to provide wireless capacity cost-effectively. Traffic hot-spots change over time, which means SCs cannot be deployed in a one-off manner as macrocells normally do, rather they should be constructed in a staged process. Hence, mathematical models that capture the time-varying staged-construction process, are urgently needed for operators to effectively predict the construction period, but are currently lacking. In this paper, inspired by the foam bursting process-a natural phenomenon that can be observed in daily life such as hand-washing, we first propose a novel model that can predict the time-varying expectation and logarithmic variance of SC coverage areas. Then, we verify the model by real network deployment cases. Additionally, in order to extract parameters from historical base station deployment data, a parameter estimation algorithm is designed and verified. The findings of the paper reveal that mobile operators should construct ultra-dense SC networks in a staged manner like how larger foams split into smaller ones.

As an important bearer network of the fifth generation (5G) mobile communication technology, the optical transport network (OTN) needs to have high-quality network performance and management capabilities. Proof by facts, the combination of artificial intelligence (AI) technology and software-defined networking (SDN) can improve significant optimization effects and management for optical transport networks. However, how to properly deploy AI in optical networks is still an open issue. The training process of AI models depends on a large amount of computing resources and training data, which undoubtedly increases the carrying burden and operating costs of the centralized network controller. With the continuous upgrading of functions and performance, small AI-based chips can be used in optical networks as on-board AI. The emergence of edge computing technology can effectively relieve the computation load of network controllers and provide high-quality AI-based networks optimization functions. In this paper, we describe an architecture called self-optimizing optical network (SOON) with cloud-edge collaboration, which introduces control-layer AI and on-board AI to achieve intelligent network management. In addition, this paper introduces several cloud-edge collaborative strategies and reviews some AI-based network optimization applications to improve the overall network performance.

Objective: This study assessed the feasibility to restore finger-specific sensory feedback in transradial amputees with electrical stimulation of evoked tactile sensation (ETS). Methods: Here we investigated primary somatosensory cortical (SI) responses of ETS using Magnetoencephalography. Results: SI activations revealed a causal correlation with peripheral stimulation of projected finger regions on the stump skin. Peak latency was accountable to neural transmission from periphery to SI. Peak intensity of SI response was proportional to the strength of peripheral stimulation, manifesting a direct neural pathway from skin receptors to SI neurons. Active regions in SI at the amputated side were consistent to the finger/hand map of homunculus, forming a mirror imaging to that of the contralateral hand. With sensory feedback, amputees can recognize a pressure at prosthetic fingers as that at the homonymous lost fingers. Conclusions: Results confirmed that the direct neural pathway from periphery to SI allows effective communication of finger-specific sensory information to these amputees.

In the fifth-generation (5G) mobile networks, the traffic is estimated to have a fast-changing and imbalance spatial-temporal distribution. It is challenging for a system-level optimisation to deal with while empirically maintaining quality of service. The 5G load balancing aims to address this problem by transferring the extra traffic from a high-load cell to its neighbouring idle cells. In recent literature, controller and machine learning algorithms are applied to assist the self-optimising and proactive schemes in drawing load balancing decisions. However, these algorithms lack the ability of forecasting upcoming high traffic demands, especially during popular events. This shortage leads to cold-start problems because of reacting to the changes in the heterogeneous dense deployment. Notably, the hotspots corresponding with skew load distribution will result in low convergence speed. To address these problems, this paper contributes to three aspects. Firstly, urban event detection is proposed to forecast the changes in cellular hotspots based on Twitter data for enabling context-awareness. Secondly, a proactive 5G load balancing strategy is simulated considering the prediction of the skewed-distributed hotspots in urban areas. Finally, we optimise this context-aware proactive load balancing strategy by forecasting the best activation time. This paper represents one of the first works to couple the real-world urban event detection with proactive load balancing.