Search

Our team had proposed a Ground Control Points-free three-dimensional interactive virtual reality (3DIVR) for building scene models. However, during the actual scene model construction, we found that the blinking-artifacts and shadow-artifacts generated by light reflection and occlusion have potential effects on corner point detection and also affect the accuracy of 3DIVR. Therefore, this paper proposes a generalized artifact elimination method for eliminating blinking-artifacts and shadow-artifacts based on previous studies. The method first identifies blinking-artifacts by detecting blinking-pixels and calculating the difference values during blinking. Then, it compares these difference values with the blinking threshold. Based on the blinking threshold, it distributes compensation values to the blinking-artifacts. Finally, it smooths the deviated data after blinking-artifact elimination. Additionally, the method eliminates shadow-artifacts by improving the local brightness balance across the three color channels. Since the most important effect of light on the image is on the brightness layer, we used a set of brightness images for a detailed implementation of the principle. And then, we used the three sets of colors to verify the effectiveness of the method proposed in this paper, comparing them with the adjusted brightness method and the adjusted background method. Compared with another method, the results obtained by the method proposed in this paper exhibit two key advantages. On the one hand, they exhibit better visual quality. On the other hand, the mean value and entropy value of our results are superior to those of the other method. Currently, our team has incorporated this method into our self-developed 3DIVR system, which, in the future, will provide better assistance to related researchers.

Attention deficit hyperactivity disorder (ADHD) is a prevalent neurodevelopmental disorder in children. Although numerous intelligent methods have been applied for its diagnosis, they seldom address symptom prediction, which is crucial for establishing the relationship between symptoms and subjective biosignals. We propose a severity prediction model, namely BSP-Net, which uses amplitude of low-frequency fluctuation (ALFF) data and constructs severity predictors within a binary hypothesis testing (BHT) framework. Specifically, we designed a dual-branch network for symptom severity prediction, with each branch corresponding to an assumed label for the test subject. Building on the accurate ADHD identification achieved by the existing auto-encoding network (AENet) within the BHT framework, we integrate its network and introduce symptom score predictors in each branch. By comparing high-level features from both branches using the AENet, we derive an estimated label. Once an assumed label is confirmed, the corresponding branch’s score predictor is activated to generate the final symptom assessment. Using the BSP-Net, our experiments achieved severity prediction accuracies of 92.4% and 84.9% on specific ADHD-200 datasets, with a score tolerance threshold of 3. Moreover, the mean squared errors on these datasets were lower than 16, significantly outperforming other methods. Importantly, discriminative brain regions corresponding to typical ALFF data in the BSP-Net were identified as ADHD biomarkers. These biomarkers align with existing research on abnormal brain regions in children with ADHD. Consequently, our method demonstrates its validity by providing biological explanations derived from ADHD mechanisms.

Road rage can easily cause serious road traffic accidents. In order to detect emotional changes of drivers in a straightforward and accurate way, and enable measures to be taken to avoid accidents, this paper presents a road rage emotion detection method called DMG-FCC for drivers of motor vehicles based on the fusion of MFCC and GFCC speech features in the frequency domain and their first-order differential features ( $\Delta $ MFCC, $\Delta $ GFCC). These features are first analysed, and each feature selected for fusion is determined preliminarily based on experience. To improve the detection efficiency, principal component analysis (PCA) is used to reduce the dimensionality of the fused features. Six principal components are selected as detection features based on the proportion of the variance values of the principal components of each dimension. The sparrow search algorithm is used to optimise the support vector machine classifier, and an SSA-SVM road rage emotion detection model is established and trained to recognise road rage emotions in drivers. The driver emotion dataset is established, and enhanced by changing the speech rate, pitch, and volume, through expansion, on the basis of the original 200 pieces of data, a total of 1400 pieces of data were got for each type of dataset. Experiments show that feature fusion gives better anti-noise performance than other single feature detection methods. Compared with other detection models, our method yields good detection performance, with a high detection accuracy of 95.2% on original dataset, and 96.8% on enhanced dataset, which can meet the real driving requirements.

Federated learning (FL) is now considered a critical method for breaking down data silos. However, data encryption can significantly increase computing time, limiting its large-scale deployment. While hardware acceleration can be an effective solution, existing research has largely focused on a single hardware type, which hinders the acceleration of FL across the various heterogeneous hardware of the participants. In light of this challenge, this paper proposes a novel FL acceleration framework that supports diverse types of hardware. Firstly, we conduct an analysis of the key elements of FL to clarify our accelerator design goals. Secondly, a unified acceleration framework is proposed, which divides FL into four layers, providing a basis for the compatibility and implementation of heterogeneous hardware acceleration. After that, based on the physical properties of three mainstream acceleration hardware, i.e., GPU, ASIC and FPGA, the architecture design of corresponding heterogeneous accelerators under the framework is detailed. Finally, we validate the effectiveness of the proposed heterogeneous hardware acceleration framework through experiments. For specific algorithms, our implementation achieves a state of the art acceleration effect compared to previous work. For the end-to-end acceleration performance, we gain $12\times $ , $7.7\times $ and $2.2\times $ improvement on GPU, ASIC and FPGA respectively, compared to CPU in large-scale vertical linear regression training tasks.

Line commutated converter based high-voltage direct-current (LCC-HVDC) transmissions are prone to harmonic oscillation under weak grids. Impedance modeling is an effective method for assessing interaction stability. Firstly, this paper proposes an improved calculation method for the DC voltage and AC currents of commutation stations to address the complex linearization of the commutation process and constructs an overall harmonic state-space (HSS) model of an LCC-HVDC. Based on the HSS model, the closed-loop AC impedances on the LCC-HVDC sending and receiving ends are then derived and verified. The impedance characteristics of the LCC-HVDC are then analyzed to provide a physical explanation for the harmonic oscillation of the system. The effects of the grid strength and control parameters on system stability are also analyzed. To improve the impedance characteristics and operating stability of the LCC-HVDC system, a virtual impedance based stability enhancement control is proposed, and a parameter design method is considered to ensure satisfactory phase margins at both the sending and receiving ends. Finally, simulation results are presented to verify the validity of the impedance model and virtual impedance based stability enhancement control.

Due to the coupling of multiple nonlinear factors in the position control system of the double closed-loop digital hydraulic cylinder, achieving precise and stable control of complex nonlinear systems with a single structural controller is difficult. To address this problem, we propose a Global Fast Terminal Sliding Mode Active Disturbance Rejection Control (GFTSMADRC). Firstly, by focusing on the mathematical model of the current double closed-loop digital hydraulic cylinder, the high-order nonlinear state equation of the system is derived, and employing active disturbance rejection control (ADRC), the high-order nonlinear system is transformed into a second-order integral series control system. Secondly, we propose Sliding Mode Active Disturbance Rejection Control (SMADRC) by integrating sliding mode variable structure control. Thirdly, to further enhance system performance, we propose a GFTSMADRC strategy by incorporating global fast terminal sliding mode variable structure control. The control law and adaptive law of the system are derived based on the Lyapunov stability theorem. Specifically, the effectiveness of the proposed control method is validated through simulation and experimentation.

Accurate carbon price prediction can help save energy and reduce emissions worldwide. Thus, this paper proposes a model that combines swarm intelligence algorithms with deep learning to predict carbon prices. In this model, we collect news related to carbon trading, construct a dictionary of carbon financial sentiment, and determine the emotional value of the carbon news. Secondly, The Harris Hawks Optimization (HHO) algorithm is improved by updating the escape energy and introducing the inertia weight. Then, the LSTM is optimized using the improved Harris Hawks Optimization (IHHO) algorithm. Finally, technical and emotional data on carbon price as multiple source input values are integrated, and the MS-IHHO-LSTM prediction model is established. The results show that the MAPE of IHHO-LSTM is 1.89%, 30.48%, and 10.30% better than that of HHO-LSTM in Hubei, Shanghai, and Shenzhen Carbon Exchanges, respectively. Similarly, MS-IHHO-LSTM showed a lower MAPE than IHHO-LSTM by 27.79%, 29.82%, and 6.33% in the corresponding regions. The results of the experiment indicate that: 1) Using IHHO to optimize LSTM hyperparameters can avoid falling into local optimal and improve prediction accuracy; 2) Incorporating emotional values can further enhance the model’s performance. The MS-IHHO-LSTM prediction model facilitates low-carbon investment, technological innovation, and green production, enabling enterprises to support environmental sustainability.

Analyzing and understanding emotional expressions in user comments is a crucial and complex task in business. Conducting text sentiment analysis is of great significance. This paper constructs a novel DIBTBL model that integrates an extended sentiment dictionary, an improved swarm intelligence algorithm, and deep learning techniques to accurately capture and analyze user emotions. Firstly, this paper expands the sentiment dictionary to extract emotion feature words from the review text. Secondly, the BERT model embeds these emotional feature words and pre-processed text into high-dimensional semantic space to obtain richer semantic representations and improve sentiment classification performance. Then, the TextCNN-BiLSTM feature extraction model is established to balance the grasping ability of local and global features. Fourthly, this paper innovatively improves the swarm intelligence algorithm BWO to optimize the parameters of the TextCNN-BiLSTM. Finally, MLP is employed for sentiment classification. The experimental data is crawled from Ctrip, China’s largest hotel booking website. In the comparative experiment, the proposed model achieves a higher accuracy than TBL, DTBL, BTBL, and IBTBL by 2.94%, 2.44%, 1.64%, and 0.66%, respectively. In addition, we compare the proposed model with seven advanced models in the open dataset waimai_10k. The experimental results indicate that this model outperforms all the other models, with an average improvement in accuracy of 8.21%. The study offers precise insights into user sentiment, assisting companies in better understanding and meeting customer needs.

With the rapid development of artificial intelligence (AI), it is foreseeable that the accuracy and efficiency of dynamic analysis for future power system will be greatly improved by the integration of dynamic simulators and AI. To explore the interaction mechanism of power system dynamic simulations and AI, a general design for AI-oriented power system dynamic simulators is proposed, which consists of a high-performance simulator with neural network supportability and flexible external and internal application programming interfaces (APIs). With the support of APIs, simulation-assisted AI and AI-assisted simulation form a comprehensive interaction mechanism between power system dynamic simulations and AI. A prototype of this design is implemented and made public based on a highly efficient electromechanical simulator. Tests of this prototype are carried out in four scenarios including sample generation, AI-based stability prediction, data-driven dynamic component modeling, and AI-aided stability control, which prove the validity, flexibility, and efficiency of the design and implementation for AI-oriented power system dynamic simulators.

Point cloud object detection is a pivotal technology in autonomous driving and robotics. Currently, the majority of cutting-edge point cloud detectors utilize Bird’s Eye View (BEV) for detection, as it allows them to take advantage of well-explored 2D detection techniques. Nevertheless, dimensionality reduction of features from 3D space to BEV space unavoidably leads to information loss, and there is a lack of research on this issue. Existing methods typically obtain BEV features by collapsing voxel or point features along the height dimension via a pooling operation or convolution, resulting in a significant decrease in geometric information. To tackle this problem, we present a new point cloud backbone network for grid-based object detection, MDRNet, which is based on adaptive dimensionality reduction and multi-level spatial residual strategies. In MDRNet, the Spatial-aware Dimensionality Reduction (SDR) is designed to dynamically concentrate on the essential components of the object during 3D-to-BEV transformation. Moreover, the Multi-level Spatial Residuals (MSR) strategy is proposed to effectively fuse multi-level spatial information in BEV feature maps. Our MDRNet can be employed on any existing grid-based object detector, resulting in a remarkable improvement in performance. Numerous experiments conducted on nuScenes, KITTI and DAIR-V have shown that MDRNet surpasses existing SOTA approaches. In particular, on the nuScenes dataset, we attained an impressive 7.2% mAP and 5.0% NDS enhancement compared with CenterPoint.

Most patients with Parkinson’s disease (PD) have different degrees of movement disorders, and effective gait analysis has a huge potential for uncovering hidden gait patterns to achieve the diagnosis of patients with PD. In this paper, the Static-Dynamic temporal networks are proposed for gait analysis. Our model involves a Static temporal pathway and a Dynamic temporal pathway. In the Static temporal pathway, the time series information of each sensor is processed independently with a parallel one-dimension convolutional neural network (1D-Convnet) to extract respective depth features. In the Dynamic temporal pathway, the stitched surface of the feet is deemed to be an irregular “image”, and the transfer of the force points at all levels on the sole is regarded as the “optical flow.” Then, the motion information of the force points at all levels is extracted by 16 parallel two-dimension convolutional neural network (2D-Convnet) independently. The results show that the Static-Dynamic temporal networks achieved better performance in gait detection of PD patients than other previous methods. Among them, the accuracy of PD diagnosis reached 96.7%, and the accuracy of severity prediction of PD reached 92.3%.

The rapid and reliable detection of oil contaminants is of crucial importance for environmental protection. This paper focused on the classification and identification of six types of oil using unconventional steady-state fluorescence spectroscopy coupled with XY-fused networks (XY-Fs). The excitation emission matrix fluorescence (EEMF) spectroscopy, synchronous fluorescence spectroscopy (SFS) and total synchronous fluorescence spectroscopy (TSFS) of the oil samples were measured by a FS920 steady-state fluorescence spectrometer. Further, the obtained results were compared with those obtained by using partial least square discriminant analysis (PLS-DA), LDA based on score matrix of multivariate curve resolution-alternating least squares (MCR-ALS-LDA) and parallel factor (PARAFAC-LDA) analysis. The experimental results showed that the best accuracy (100%) for the test samples was obtained by TSFS combined with XY-Fs. These results provide important references for discrimination of oil contaminants.

To eliminate inter-symbol interferences (ISI) caused by nonlinear group delay effect in high-speed broadband satellite sensor networks, an efficient parallel adaptive blind equalization of improved block constant modulus algorithm (IBCMA) with decision-directed (DD) mode (IBCMA-DD) is proposed with block minimum mean square error (MMSE) criterion. First, nonlinear group delay models are introduced in satellite channels. Second, the criterion of the classical Godard’s constant modulus algorithm (CMA) is derived in the block vector form as the IBCMA for satellite signals with inter-symbol interference (ISI) caused by the above satellite channels. In the IBCMA, stochastic gradient descent and block minimum mean square error are adopted to reduce excess errors in equalization by adjusting block tap-weight vectors adaptively. Finally, the IBCMA is combined with the DD mode of equalization to both accelerate the parallel blind equalization with low complexity and improve the spectrum efficiency without any training data to occupy unnecessary bandwidth. Simulation results indicate that the proposed IBCMA-DD equalization possesses good performance in approaching ideal non-ISI transmissions. It achieves good bit error rate (BER) performance just within 0.3 dB of the theoretic performance under high signal-to-noise ratios (SNRs), even at satellite group delay channels by high-power nonlinear broadband filters. Therefore, it obtains excellent equalization under severely deteriorated group delay channels in broadband satellite sensor networks with high data rate.

Soft rehabilitation exoskeletons have gained much attention in recent years, striving to assist the paralyzed individuals restore motor functions. However, it is a challenge to promote human-robot interaction property and satisfy personalized training requirements. This article proposes a soft elbow rehabilitation exoskeleton for the multi-mode training of disabled patients. An adaptive cooperative admittance backstepping control strategy combined with surface electromyography (sEMG)-based joint torque estimation and neural network compensation is developed, which can induce the active participation of patients and guarantee the accomplishment and safety of training. The proposed control scheme can be transformed into four rehabilitation training modes to optimize the cooperative training performance. Experimental studies involving four healthy subjects and four paralyzed subjects are carried out. The average root mean square error and peak error in trajectory tracking test are 3.18° and 5.68°. The active cooperation level can be adjusted via admittance model, ranging from 4.51 °/Nm to 10.99 °/Nm. In cooperative training test, the average training mode value and effort score of healthy subjects (i.e., 1.58 and 1.50) are lower than those of paralyzed subjects (i.e., 2.42 and 3.38), while the average smoothness score and stability score of healthy subjects (i.e., 3.25 and 3.42) are higher than those of paralyzed subjects (i.e., 1.67 and 1.71). The experimental results verify the superiority of proposed control strategy in improving position control performance and satisfying the training requirements of the patients with different hemiplegia degrees and training objectives.

A feature-tracking algorithm algorithm utilizing the proposed K-nearest-neighbor vector field consensus (KVFC) to filter outliers is developed to monitor the dynamic changes of sea ice retrieval from synthetic aperture radar (SAR) images. The KVFC is based on vector field consensus and combines with local neighborhood correspondences to optimize the elimination of outliers while retaining inliers as many as possible. The proposed KVFC was evaluated and compared with several algorithms on three standard datasets and Sentinel-1 image pairs in the Fram Strait and the Beaufort Sea. The KVFC obtained more sea-ice drift vectors than the nearest neighbor similarity ratio (NNSR) with a 0.7 threshold and generated dense distribution of sea-ice drift vectors combined with the HH and HV channels. Using buoy datasets to calculate the sea-ice drift speed and evaluate algorithm performance, the proposed approach yielded a lower mean error (KVFC: −0.150 cm/s, NNSR: 0.407 cm/s), lower root mean square error (KVFC: 0.476 cm/s, NNSR: 1.817 cm/s), and lower angle deviation (KVFC: 3.542°, NNSR: 10.318°) compared to the NNSR.

Federated learning has solved the problems of data silos and data fragmentation on the premise of satisfying privacy. However, cryptographic algorithms in federated learning brought significant increase in computational complexity, which limited the speed of model training. In this paper, we propose a hardware/software (HW/SW) co-designed field programmable gate array (FPGA) accelerator for federated learning. Firstly, we analyzed the time consumption of each stage in federated learning and the involved cryptographic algorithms, and found the performance bottleneck. Secondly, a HW/SW co-designed architecture is introduced, which can speed up encryption, decryption and ciphertext-space computation at the same time without reconfiguring FPGA circuit. In the HW part, we proposed a Hardware-aware Montgomery Algorithm (HWMA) which utilized data parallelism and pipeline, and designed an FPGA architecture to decouple data access and computation. In the SW part, an Operator Scheduling Engine (OSE) is designed, which can flexibly resolve the target algorithm into multiple HWMA calls, and complete other non-computation-intensive calculations. Finally, evaluations for both specific algorithms and practical applications are implemented. Experimental results show that when deployed on Intel Stratix 10 FPGA, our accelerator can increase the throughput of 2048-bit modular multiplication, modular exponentiation and Paillier algorithm to more than 3x of the CPU. When integrated into a industrial grade federated learning open source framework, the end-to-end training time of linear regression and logistic regression can be shortened by 2.28x and 3.30x respectively, which is more than 2x faster than the reported best results of FPGA accelerator.

The moving image deblurring method based on deep learning has achieved good results. However, some methods are not effective in restoring image texture detail information. Therefore, this paper proposes a High-Frequency Attention Residual Module (HFAR), which is used to guide the network to learn more high-frequency texture information in the image to improve the quality of image detail restoration. The designed attention residual module consists of two sub-modules, Fourier Channel Attention module (FCA) and Edge Spatial Attention module (ESA). The FCA module gives more weight to the feature maps that contain more high-frequency information in multiple channels. While the ESA module gives more weight to the areas in the feature maps which contain more high-frequency information to guide the network to learn image details and texture information. Extensive experiments on different datasets show that our method achieves state-of-the-art performance in motion deblurring.

It is well known that the air-gap magnetic field of the permanent magnet synchronous generator (PMSG) is difficult to adjust. To overcome this challenge while maintaining the brushless feature of PMSGs, a hybrid excitation synchronous generator (HESG) based on inherent third harmonic field induction is proposed in this paper. The HESG adds a set of third harmonic windings with a pitch of 1/3 of the armature winding in the stator core and a set of harmonic windings with the same pitch as the third harmonic winding in the rotor core. Due to electromagnetic induction, the third harmonic field established by third harmonic windings in the stator induces the harmonic electromotive force (EMF) of six times the frequency in the rotor harmonic winding. This harmonic EMF is directly provided to the rotor excitation winding after rectifying by the diode rotating rectifier. The HESG has armature windings and third harmonic windings in the stator, and excitation windings, harmonic windings and permanent magnets in the rotor. In this article, the mechanism of stator third harmonic winding EMF, rotor harmonic EMF, and the brushless excitation are elaborated. The experimental results of adjustable armature winding voltages with controlling the stator third harmonic winding current have proven the feasibility of the proposed brushless excitation principle based on the third harmonic.

Monitoring the depth of unconsciousness during anesthesia is beneficial in both clinical settings and neuroscience investigations to understand brain mechanisms. Electroencephalogram (EEG) has been used as an objective means of characterizing brain altered arousal and/or cognition states induced by anesthetics in real-time. Different general anesthetics affect cerebral electrical activities in different ways. However, the performance of conventional machine learning models on EEG data is unsatisfactory due to the low Signal to Noise Ratio (SNR) in the EEG signals, especially in the office-based anesthesia EEG setting. Deep learning models have been used widely in the field of Brain Computer Interface (BCI) to perform classification and pattern recognition tasks due to their capability of good generalization and handling noises. Compared to other BCI applications, where deep learning has demonstrated encouraging results, the deep learning approach for classifying different brain consciousness states under anesthesia has been much less investigated. In this paper, we propose a new framework based on meta-learning using deep neural networks, named Anes-MetaNet, to classify brain states under anesthetics. The Anes-MetaNet is composed of Convolutional Neural Networks (CNN) to extract power spectrum features, and a time consequence model based on Long Short-Term Memory (LSTM) networks to capture the temporal dependencies, and a meta-learning framework to handle large cross-subject variability. We use a multi-stage training paradigm to improve the performance, which is justified by visualizing the high-level feature mapping. Experiments on the office-based anesthesia EEG dataset demonstrate the effectiveness of our proposed Anes-MetaNet by comparison of existing methods.

The applications of super-resolution (SR) technology in the field of image completion are successful. Nevertheless, industry applications demand not only image completion but also the topology and time-series completion. In this article, the SR technology on a topology graph is studied in the scenario of recovering measurements in power distribution systems for cost saving and security & stability improvement. The power flow and voltage magnitude measurements on feeders are reported at different frequencies. In this article, a new data completion method considering distribution system topology is proposed. Firstly, the graph convolutional neural network (GCN) is used for spatial-temporal convolution on a graph, and then the power system state estimation (SE) is used introducing the physical constraints. This method realizes the super-resolution of distribution system measurements, improves the state awareness of distribution systems. Hence, it helps to improve the efficiency of distribution network operation and to reduce equipment failures.

In this paper, we solve the problem of dynamic scenes deblurring with motion blur. Restoration of images in the presence of motion blur necessitates a network design that the receptive field can completely cover all areas that need to be deblurred, while the existing network increases the receptive field by continuously stacking the ordinary convolutional layer or increasing the size of the convolution kernel. However, these methods inevitably increase the computational burden of the network. We propose a novel architecture consisting of a channel adaptive residual module. Different features of the blurred image are extracted and distributed on each feature channel. Our network can calculate the weight of each channel through learning, and extract the image features adaptively according to different degrees of blurring and importance of information. We embed the module in a modified encoder-decoder design with skip connections to achieve multi-scale feature fusion for further performance improvement. The extensive comparison with the existing techniques in the baseline dynamic scene deblurring dataset shows that the proposed network can effectively realize image deblurring, and the accuracy and speed are comparable with the existing techniques.

Lung nodule classification plays an important role in diagnosis of lung cancer which is essential to patients’ survival. However, because the number of lung CT images in current dataset is relatively small and the ratio of nodule samples to non-nodule samples is usually very different, this makes the training of neural networks difficult and poor performance of neural networks. Hence, LDNNET is proposed, which adopts Dense-Block, batch normalization (BN) and dropout to cope with these problems. Meanwhile, LDNNET is an adaptive architecture based on convnets combining softmax classifier which is utilized to alleviate the problems of training deep convnets. Follows are our main work: Firstly, we utilized LDNNET on database LUng Nodule Analysis 2016 (LUNA16) for lung nodule classification and database KAGGLE DATA-SCIENCE-BOWL-2017(Kaggle DSB 2017) for lung cancer classification; Secondly, the comparison experiments are designed to compare the performance of dense connection, pooling layer and the input pixel size of lung CT(Computed Tomography) images; Thirdly, data enhancement, dense connection and dropout layer were utilized in LDNNET to reduce overfitting; Fourthly, pre-processing methods, for instance enhanced contrast, median filtering, Laplacian filtering are compared to the no-processing method to explore the effect of pre-processing on lung CT images classification. Fifthly, accuracy, specificity and sensitivity on LUNA16 are 0.988396, 0.994585 and 0.982072 and these indicators on Kaggle DSB 2017 are 0.999480, 0.999652 and 0.998974. Furthermore, AUC for both two datasets is over 0.98. Consequently, this paper conducts experiments with uniform parameter settings on two publicly available databases and shows that even in challenging situation where lung images are directly utilized as input images without preprocessing, LDNNET is still the more advanced algorithm than other recent algorithms respectively. Moreover, a series of comparative experiments were conducted to further confirm that the proposed algorithm has the higher accuracy and robustness through verification and discussion.

In the intensive care unit(ICU), excessive false alarms burden medical staff greatly, and cause a waste of medical resources as well. In order to alleviate false alarming in ICU, we constructed classification models based on convolutional neural network, which can directly deal with time series and avoid manually extracting features. Combining with grouping strategy, we tried two basic network structures, i.e. deep group convolutional neural network(DGCN) and embedded deep group convolutional network(EDGCN). After that, based on EDGCN, which was proved better, ensemble networks were constructed to elevate the performance further. Considering of the limited sample size, various data expansions were also tried. For comparison, we used a widely adopted index ‘Score’, which is defined as $100\times $ (TP+TN)/(TP+TN+FP $+ 5\times $ FN), as performance evaluation. Finally, we tested our model in the online sandbox, and got a score of 80.68. Although it is slightly lower than the best scores that have been reported, our models are end-to-end, which means the original time series can be automatically mapped into a binary output, without manual feature extraction. In addition, the innovative adoption of group convolution makes full use of information from multi-channel signals. Besides, this work provides ideas for relating network parameter selection with physiological or physical natures of signals. In the end, after further discussion, we also proposed potential elevation solutions.

The papers in this special section focus on applied artificial inteligence and machine learning for video coding and media streaming.

This paper proposes a novel siamese cross contrast neural network (SCCNN) to classify the hepatocellular carcinoma (HCC) and the intrahepatic cholangiocarcinoma (ICC) on computed tomography (CT) images. This method is inspired from cross contrast neural networks (CCNN) which is based on tailored CNN and information based similarity(IBS) theory. A new IBS-based measurement named as discriminative IBS(DisIBS) is designed for SCCNN. SCCNN is composed of two main parts including siamese feature extractors with DisIBS operator and MLP classifiers. Siamese networks extract features with DisIBS calculated by DisIBS operator as metric at the top. MLP classifiers are connected with but gradient-stop to feature extractors deriving classification results. We assign different loss functions with different parts to make better practice, specially DisIBS-based loss for feature extractors and softmax-based for MLP classifiers. SCCNN preserves the advantages of CCNN that can fit the insufficient medical images and small lesions. Furthermore, it extends CCNN with the siamese mechanism and gradient-stop MLP classifiers to accept the random inputs and predict like traditional CNN. To present the effectiveness of SCCNN empirically, we apply this method on a 234-person (157/77 for train/test) dataset and achieve better results than other classic CNN and CCNN methods. We try different base models of siamese structures and display prediction accuracy in two levels (slice/patient). The highest slice/patient accuracy which we have achieved on three-categories classification (HCC/ICC/Normal) is 90.22%/94.92% and the accuracy rises to 94.17%/97.44% on binary classification(HCC/ICC).

Attention Deficit Hyperactivity Disorder (ADHD) is a high incidence of neurobehavioral disease in school-age children. Its neurobiological diagnosis (or classification) is meaningful for clinicians to give proper treatment for ADHD patients. The existing ADHD classification methods suffer from two problems, i.e., insufficient data and noise disturbance. In this paper, a high-accuracy classification method is proposed by using brain Functional Connectivity (FC) as ADHD features, where an l2,1-norm Linear Discriminant Analysis (LDA) model and a binary hypothesis testing framework are effectively employed. In detail, we introduce a binary hypothesis testing framework to cope with insufficient data of ADHD database. The FCs of test data (without seeing its label) are used for training and thus affect the subspace learning of training data under binary hypotheses. On other hand, the l2,1-norm LDA model generates a subspace to represent ADHD features, aiming to overcome noise disturbance. By robustly learning ADHD features, the subspace energy difference between binary hypotheses becomes more discriminative. Thereby, the true hypothesis can be rightly estimated with its larger subspace energy, which provides reliable evidence to predict the label of test data. By the test on ADHD-200 database, it shows that our method outperforms other state-of-the-art methods with the significant average accuracy of 97.6%. Moreover, the corresponding result analysis with ADHD symptom score and the explanation of discriminative FCs between ADHD and healthy control groups are given, which further verifies the validity of our classification method.

Extreme disasters may cause the power supply to the distribution system (DS) to be interrupted. The DS is forced to operate in island mode and forms an islanded microgrid (MG). In order to improve the post-disaster resilience of the DS and to provide longer power supply for as many loads as possible with limited generation resources, this paper proposes a multi-agent deep reinforcement learning (DRL) method which realizes a dual control on the source and load sides of the MG. The problem of resilience improvement is converted to a sequential decision making problem, where the objective is to maximize the cumulative MG utility value over the power outage duration. A multi-agent DRL model is proposed to solve the sequential decision making problem. A dual control policy including energy storage management and load shedding strategy is put forward to maximize the utility value of the MG. A reinforcement learning (RL) environment based on OpenAI and OpenDSS for islanded MG is constructed as a simulator, which has a general interface compatible with, and also can be published to, OpenAI Gym. Numerical simulations are performed for an MG equipped with wind turbines, diesel generators, and storage devices to validate the effectiveness of the proposed method. The influences of available generation resources and power outage duration on the control policy are discussed, which validates the strong adaptability of the proposed method in different conditions.

This paper presents a novel method to identify a class of closed-loop systems, in which both the forward channel and the feedback channel have unknown time-delays. Taking into account the time-delays, an overparameterized identification model with a sparse parameter vector is established. Based on the basis pursuit de-noising criterion, the sparse parameter vector is estimated by solving a quadratic programming. The time-delays and the parameters are estimated according to the structure of the parameter estimation vector and the model equivalence principle, respectively. The proposed method is applicable even in the case of a few number of sampled data. The effectiveness of the proposed algorithm is verified by the numerical simulation results.

Lung parenchyma segmentation is the prerequisite for an automatic diagnosis system to analyze lung CT (computed tomography) images. However, traditional lung segmentation algorithms have poor adaptability and are not effectively robust regarding lung databases with blood vessels and small voids which can interfere the segmentation. The main work of this paper is as follows: Firstly, a lung dense deep convolutional neural network (LDDNet) is proposed, which adopts some popular optimizer methods, such as dense block, batch normalization (BN) and dropout. The performance of LDDNet is tested on the public lung database LIDC-IDRI which contains many cases of interference for segmentation. Secondly, the labeled with blood vessels and small voids are not contained by the public ground-truth masks of the LIDC-IDRI database, therefore these regions are labeled by us with LabelMe software. Thirdly, for the aim of exploring the effect of image preprocessing on segmenting lung CT images with deep neural network, contrast enhancing, median filtering and Laplacian filtering are used to preprocess the image as comparative experiments. Finally, dataset is classified into four classes by the geometrical shapes to test the performance of LDDNet. The accuracy of the segmentation experiment reaches over 99% and the four classes can all reach over 95%. Additionally, blood vessels and small voids are segmented out from the lung parenchyma which is not achieved by other methods. Experimental results confirm that the proposed LDDNet can segment the lung parenchymal area more accurately and has better robustness in comparison with other neural networks and most of the traditional methods.

Voice Conversion (VC) is a method of converting the source speaker's speech into the target speaker's speech without changing the source speaker's speech content. The current VC methods have the following problems: (1) they are only applicable to a limited number of speakers, not to any speakers, as a result, the application scenarios are greatly restricted; (2) the representation (feature) separation(RS) effect of the current mainstream technology is not ideal on the source speaker speech and the target speaker speech; and (3) the voice conversion quality of most models is unsatisfactory, and hence needs to be improved. Therefore, in this paper, we constructed a one-shot VC model of Representation Separation, called RS-VC model, implemented by the encoder-decoder structure. The encoder is composed of a content encoder and a speaker encoder. The content encoder separates the content information of the source speaker voice and generates a content representation. The speaker encoder separates the target speaker information of the target speaker voice and generates a speaker representation. The decoder synthesizes the content representation and the speaker representation to generate the converted voice. In this paper, we obtained the optimized speaker verification model SVIGEN2E (Speaker Verification with Instance Normalization using Generalized End-to-End loss) by improving the speaker verification (SV) model. The model SVIGEN2E is used as the speaker encoder. This speaker encoder needs to be trained in advance prior to RS-VC model training, and the pre-trained model of SVINGE2E directly extracts speaker representation of the target speaker's voice, and is used for training and testing RS-VC model. A progressive training method is proposed then for training RS-VC model. Experiments show that the progressive training method can effectively improve the quality of the converted voice. Compared with the basic speaker verification model, both SVINGE2E and RS-VC deliver the impressive improvements in EER (Equal Error Rate).

Real-time accurate detection of three-dimensional (3D) objects is a fundamental necessity for self-driving vehicles. Most existing computer vision approaches are based on convolutional neural networks (CNNs). Although the CNN-based approaches can achieve high detection accuracy, their high energy consumption is a severe drawback. To resolve this problem, novel energy efficient approaches should be explored. Spiking neural network (SNN) is a promising candidate because it has orders-of-magnitude lower energy consumption than CNN. Unfortunately, the studying of SNN has been limited in small networks only. The application of SNN for large 3D object detection networks has remain largely open. In this paper, we integrate spiking convolutional neural network (SCNN) with temporal coding into the YOLOv2 architecture for real-time object detection. To take the advantage of spiking signals, we develop a novel data preprocessing layer that translates 3D point-cloud data into spike time data. We propose an analog circuit to implement the non-leaky integrate and fire neuron used in our SCNN, from which the energy consumption of each spike is estimated. Moreover, we present a method to calculate the network sparsity and the energy consumption of the overall network. Extensive experiments have been conducted based on the KITTI dataset, which show that the proposed network can reach competitive detection accuracy as existing approaches, yet with much lower average energy consumption. If implemented in dedicated hardware, our network could have a mean sparsity of 56.24% and extremely low total energy consumption of 0.247mJ only. Implemented in NVIDIA GTX 1080i GPU, we can achieve 35.7 fps frame rate, high enough for real-time object detection.

Technology determines tactics, and tactics promote the development of technology. Artificial intelligence technology is a multiplier that accelerates the innovation and development of military theory. In this paper, we first list the different intelligence levels by introducing their corresponding applications. Then we review the technical classification based on the related concepts, Finally, we discuss technical and practical difficulties and give some solutions from the aspects of strengthening knowledge engineering, building simulation systems, and accumulating data engineering knowledge. The development of Artificial intelligence technology has a profound impact on military development trends, leading to major changes in the forms and modes of war.

As wireless communication technology and semiconductor technology developed fast, the wireless medical sensor networks (WMSNs) have been applied to the modern health-care area at large. The physiological data can be obtained by medical sensor nodes deployed in the patient's body and sent to the special devices of the health professionals through wireless communication. Thus, the status of a patient is monitored by the health professional in that way. However, there are still two important issues that how to guarantee secure communication and protect the privacy of the patient for the reason of the open feature of wireless communication. In this paper, initially, an improved three-factor user authentication scheme is proposed to overcome those flaws utilizing password, smart card, and biometric feature. Furthermore, formal security analysis shows that the proposed scheme defends against various security pitfalls. Finally, the comparison results with other surviving relevant schemes show that our scheme is more efficient in terms of computational cost, communication cost, and estimated time. Therefore, the proposed scheme is suitable for practical application in WMSN.

In this paper, we propose a failure mechanism cumulative model that considers the loading history of the load sharing effect in a k-out-of-n system. Three types of failure mechanisms are considered, continuous degradation, compound point degradation, and sudden failure due to shock. By constructing a logic diagram with functional dependence gate, the load-sharing effect can be explained from the failure mechanism point of view using a mechanism-acceleration gate that shows that when one component fails, the failure mechanisms of the other surviving components will be accelerated. By deriving the total damage equation and a constructing failure behavior model, the system reliability of a k-out-of-n system with different types of failure mechanisms were evaluated. A voltage stabilizing system that contains a 1-out-of-2 subsystem, or a 2-out-of 3 subsystem was used to illustrate the practical applicability of the proposed approach. A combined Monte Carlo and binary decision diagram method was used in the numerical simulation process.

Failure scenarios, which form the basis for accident scenarios, need to be studied to describe the failure behavior of complex systems. This paper proposes a hybrid intelligent method that combines the A* intelligent algorithm with the breadth-first search algorithm to automatically generate the failure scenario of a complex system with the failure scenario tree, while simultaneously calculating the occurrence probability of each failure path and of the whole system. The simulation is guided by the failure behavior rules generated based on expert knowledge. A case study of a power supply system with a warm standby subsystem is conducted. This system is also a multi-state system. The obtained results show that the proposed automatic reasoning can identify key failure scenarios that induce system failure, which can be helpful in the decision-making process.

Distributed state estimation (DSE) is effective for the analysis of multi-region state estimation problems, avoiding integrating complicated data sets and models. In this work, the peer-to-peer decomposition of the Karush-Kuhn-Tucker (KKT) condition is applied to formulate the fully distributed fast decoupled state estimation (FD2SE) algorithm. Then, a fixed-matrix form data-exchanging interface incorporating the fast decoupled algorithm is designed, reducing the communication costs and allowing it to be used for various system operating conditions. Further, a convergence detecting method for dispatch centers is proposed in order to infer, locally and asynchronously, about the global convergence. Case studies carried out on IEEE systems and a real power grid of China verify the accuracy and efficiency of the FD2SE algorithm.

Iris segmentation algorithms are of great significance in complete iris recognition systems, and directly affect the iris verification and recognition results. However, the conventional iris segmentation algorithms have poor adaptability and are not sufficiently robust when applied to noisy iris databases captured under unconstrained conditions. In addition, there are currently no large iris databases; thus, the iris segmentation algorithms cannot maximize the benefits of convolutional neural networks (CNNs). The main work of this paper is as follows: first, we propose an architecture based on CNNs combined with dense blocks for iris segmentation, referred to as a dense-fully convolutional network (DFCN), and adopt some popular optimizer methods, such as batch normalization (BN) and dropout. Second, because the public ground-truth masks of the CASIA-Interval-v4 and IITD iris databases do not include the labeled eyelash regions, we label these regions that occlude the iris regions using the Labelme software package. Finally, the promising results of experiments based on the CASIA-Interval-v4, IITD, and UBIRIS.V2 iris databases captured under different conditions reveal that the iris segmentation network proposed in this paper outperforms all of the conventional and most of the CNN-based iris segmentation algorithms with which we compared our algorithm's results in terms of various metrics, including the accuracy, precision, recall, f1 score, and nice1 and nice2 error scores, reflecting the robustness of our proposed network.

Low-power wide-area (LPWA) communication has gained increasing attention in recent years with the rapid growth of fifth generation evolution (5G), the Internet of Things (IoT), and mobile computing. Narrowband Internet of Things (NB-IoT) is one kind of LPWA technology based on cellular IoT, which supports massive connections, wide area coverage, ultra-low power consumption, and ultra-low cost. Research on NB-IoT communication is increasingly attractive. Network calculus theory facilitates the performance analysis and network optimization of the NB-IoT system. We aim to analyze and optimize the NB-IoT networks. In this paper, we construct a random access traffic model including the NB-IoT user equipment (UE) arrival process and eNB service process. Then, we utilize the stochastic network calculus (SNC) to analyze the network delay in NB-IoT traffic model. Random latency bounds in different arrival processes are derived. Simulations show that SNC can evaluate the system delay under different distributions effectively. For the condition that numerous UEs access simultaneously following the Beta distribution, we first propose an improved K-means algorithm to cluster the NB-IoT terminals. Then, we raise the scheduling strategy on the basis of priority. It consists of the priority generation algorithm IPGNTQ and the NB-IoT task scheduling algorithm SANTQ. The extensive experiment results verify that our proposed optimized strategy can alleviate the network congestion effectually. Moreover, we compare our proposed optimized scheme with four existing uplink traffic scheduling schemes, showing that ours outperforms all of them.

Through virus-based reconnaissances, the attacker may compromise critical power substations and induce power outages. Meanwhile, the power system operator may conduct defensive measures to prevent virus-spreading. In this paper, a graphical evolutionary game model is developed to solve the equilibrium of competitions between virus propagations and countermeasures, which helps to evaluate the cyber-security risk of power systems under long-term virus-based intrusions. In the proposed game model, the virus is supposed to propagate itself in a cyber network connecting cyber nodes belonging to power substations. Each cyber node may play as an attacker or a defender according to its state. A death-birth rule is adopted to model probabilistic strategies and state transfers, which reflects the bounded rationality of the viruses. Then, an evolution dynamic equation is derived to obtain the portion of the infected substations when the equilibrium is reached, and simulation algorithm is also given to estimate the infection probability of each substation under the persistent virus-based intrusion. Assuming a virus may trip all breakers in the infected substation, the cyber-security risk is evaluated to identify system-level vulnerabilities. The IEEE 118-bus system with a layered cyber network is used for case studies. Test results validate the proposed model and provide further insights into the virus-based intrusion.

Driven by the vision of 5G communication, the demand for mobile communication services has increased explosively. Ultra-dense networks (UDN) is a key technology in 5G. The combination of mobile edge computing (MEC) and UDN can not only cope with access from mass communication devices, but also provide powerful computing capacity for users at the edge of wireless networks. The UDN based on MEC can effectively process computation-intensive and data-intensive tasks. However, when a large number of users offload tasks to the edge server, both the network load and transmission interference would increase. In this paper, the problem of task offloading and channel resource allocation based on MEC in 5G UDN is studied. Specifically, we formulate task offloading as an integer nonlinear programming problem. Due to the coupling of decision variables, we propose an efficient task offloading and channel resource allocation scheme based on differential evolution algorithm. Simulation results show that the proposed scheme can obviously reduce energy consumption and has good convergence.

Heave compensation systems are widely used to decouple the load motion from wave-induced vessel motion for the equipment handling on the ocean. Researches have been made to achieve successful compensation, yet few of them discusses the inherent constraints of the systems, such as bounded compensator's stroke and max actuator's velocity. This paper presents a solution for active heave compensation systems with such constraints by means of variable structure control. The controller's complexity on design procedures and effectiveness are compared with a trajectory planning control method which turns out that the variable structure controller is more suitable to apply to the active heave compensators. The back-stepping method is used to robustly stabilize this variable structure system and for the aim of a decrease on the high robust gain due to uncertain friction term, a modified decoupled friction observer is used which is also verified by both theoretical and experimental analyses. To compensate for the time delay of the motion reference unit (MRU), a heave prediction algorithm is used. The experimental results show that most heave motion can be compensated when the motion and its velocity are feasible, while no hit occurs otherwise.

User-centric network (UCN) organizes a dynamic base station group (BSG) for each user equipment (UE), consisting of potential base stations (BSs) with caching ability. However, the performance gain brought by the increased BS diversity of potential BSs has not yet been analyzed. Using stochastic geometry, the analytical expression of average outage probability is derived for the cache-enabled UCN, where a practical scenario is considered wherein inter-cache data transfer cannot be performed between BSs and the serving BS must have the requested content. Specifically, the gain of BS diversity within the BSG is characterized, where the UE benefits from transmission resource diversity of the BSG by always connecting to the BS with the best coverage performance among those with the required content, and the increased caching content diversity is utilized by searching the content within the BSG. Based on the analysis, to minimize the lower bound of outage probability, closed-form optimal caching distribution is obtained which is more uniform than the request distribution. Outage probability of the cache-enabled UCN is decreased by 42.1% compared to the case where the UE connects to the nearest BS with the required content. Moreover, the optimal caching distribution decreases outage probability by 28.5% compared to the policy of caching the most popular contents.

Power systems used on shipboards, airplanes, and similar applications are usually classified as isolated power systems (IPSs), where pulsed loads are very important under both normal and extreme conditions. Because of the high energy density of pulsed loads, stable operation of an IPS with multiple types of pulsed loads faces great challenges. This paper proposes an optimal operation planning of IPS with multiple pulsed loads while considering stability constraints. To mitigate the negative effect of pulsed loads, energy storage devices are used to buffer the required energy. The stability constraints are derived from a proposed analytical criterion whose feasibility is guaranteed by fully studied principles. A critical point-based particle swarm optimization algorithm is proposed to solve the optimal planning problem. The case studies on a typical IPS with multiple pulsed loads verify the correctness and effectiveness of the proposed method.

Complicated systems are always required to perform different missions each of which may include several phases. Meanwhile, these systems and their components can get exhausted and deteriorated during the operating processes, which drives systems and components into different states. These systems, which is in possession of the characteristics of multiple states and phased missions simultaneously, are called multi-state and phased-mission systems (MS-PMSs). During the last decade, some scholars have made considerable efforts to study the analyzing methods to the reliability of MS-PMSs. However, compared with the independent researches into multi-state systems or phased-mission systems, the achievement of MS-PMSs is still insufficient. Besides, these studies focused on the modeling methods to the MS-PMSs from the system level and the failure behavior was seldom mentioned. In this paper, by extending our previous studies, two accumulating rules of failure mechanism are provided and a revised hierarchical BDD-based method is proposed to model and analyze the reliability of MS-PMSs. Finally, a simplified realistic example is given to illustrate the using of the proposed method.

Electromagnetic transients (EMT) simulation is the most accurate and intensive computation for power systems. Past research has shown the potential of accelerating such simulations using graphics processing units (GPUs). In this paper, an efficient GPU-based parallel EMT simulator is designed. Threadoriented model transformations are first proposed for the electrical and control systems. Following the transformations, the electrical system is represented by connected networks of massive primitive electrical elements, the computations of which can be constructed as massive fused multiply-add operations and solutions to a linear equation. The control systems are represented by a layered directed acyclic graph with primitive control elements that can be dealt with using single-instruction-multiple-threads groups. Finally, code automation tools are designed to form the GPU kernels. Compared with past work, the proposed model transformations improve the degree of parallelism. Most importantly, the code automation tools improve computational efficiency by substantially reducing addressing and memory access, and render the implementation of the algorithm more general and convenient. Test systems of different sizes were created by connecting multiple IEEE 33-bus distribution systems and adding distributed generators. Simulations were performed on NVIDIA's K20x and P100 cards. The results indicate that the proposed method significantly accelerates EMT simulations compared with a CPU-based program. Real-time performance was also achieved under certain conditions.

Changes in motor behavior during aging might be induced by complexity fluctuations in the neuromuscular system. Most previous studies have been performed based on single-scale entropies. In this paper, multiscale fuzzy entropy (MSFuzzyEn) was applied to characterize the changes in the complexity of simulated electromyogram (EMG) signals with the increasing motor unit number and signal-to-noise ratio. Age-related differences in multiscale complexity during handgrip control were also investigated. Ten young and 10 older adults were instructed to produce constant forces at 25%, 50%, and 75% of their maximal grip force with their dominant hands. The grip force and EMG signals of four forearm muscles were recorded simultaneously and analyzed using MSFuzzyEn. The simulation tests revealed that, as the time scale increased, the interference of noise in the EMG signals decreased. At time scale 1, the complexities of the force and EMG signals exhibited opposite changes with aging. When the time scale increased, we observed a loss in complexity with aging in both the force and EMG signals. These results confirmed the merits of MSFuzzyEn in noise abatement, and implied that entropy at relatively larger time scales might better characterize EMG signals. Further studies should extend the application of multiscale entropy in pathologies.

The false data injection (FDI) attack is a potential threat to the security of smart grids, and therefore, such threats should be assessed carefully. This paper proposes a self-governing FDI attack method with exploitation and exploration mechanisms and then evaluates its threat to power systems. The attack is executed by viruses embedded in monitoring systems in substations. First, the FDI attack is formulated as a partially observable Markov decision process. Then, an improved online reinforcement learning method called nearest sequence memory Q-learning is adopted to make the attack more effective. Finally, propagation, an inherent property of viruses, is described using a propagation-evolution model that serves as the exploration mechanism for the proposed FDI attack. To validate the proposed attack method, cosimulations of daily operations of the IEEE 39-bus system are performed in which both the automatic voltage control system and the proposed FDI attack are modeled. Test results show that the proposed FDI method can cause voltage collapse even if only a few substations are infected.

Abstract In a hybrid AC/DC medium voltage distribution network, distributed generations (DGs), energy storage systems (ESSs), and the voltage source converters (VSCs) between AC and DC lines, have the ability to regulate node voltages in real-time. However, the voltage regulation abilities of above devices are limited by their ratings. And the voltage regulation efficiencies of these devices are also different. Besides, due to high r/x ratio, node voltages are influenced by both real and reactive power. In order to achieve the coordinated voltage regulation in a hybrid AC/DC distribution network, a priority-based real-time control strategy is proposed based on the voltage control effect of real and reactive power adjustment. The equivalence of real and reactive power adjustment on voltage control is considered in control area partition optimization, in which regulation efficiency and capability are taken as objectives. In order to accommodate more DGs, the coordination of controllable devices is achieved according to voltage sensitivities. Simulations studies are performed to verify the proposed method.

Overlapped frequency division multiplexing (OVFDM) systems can obtain high spectral efficiency (SE), which is proportional to the constraint length. However, high decoding complexity imposes the main challenge on OVFDM systems. This paper proposes a low-complexity sliding window (SW) block decoding algorithm for OVFDM systems, where data symbols are estimated based on the reception of a SW instead of a date frame. Specifically, block code of each SW is decoded by bit-flipping algorithm where the bits to be flipped are selected according to the largest absolute value criterion. Using this criterion, the complexity to obtain the near optimal bit-flipping vector grows only linearly with the SW length. In addition, the study of the decoding algorithm is based on the design of OVFDM encoding structure, where symbols can occupy orthogonal in-phase and quadrature channels simultaneously to further improve SE by a factor of two. Simulation results show that OVFDM SW decoding with bit-flipping algorithm can be used when constraint length is relatively high (constraint length ≥ 20) because the complexity goes roughly linearly with the increase of constraint length.

In this paper, the structural controllability of the systems over F(z) is studied using a new mathematical method-matroids. First, a vector matroid is defined over F(z). Second, the full rank conditions of [sI - A|B](s ∈ p) are derived in terms of the concept related to matroid theory, such as rank, base, and union. Then, the sufficient condition for the linear system and composite system over F(z) to be structurally controllable is obtained. Finally, this paper gives several examples to demonstrate that the married-theoretic approach is simpler than other existing approaches.