最新刊期
卷 47 ， 期 8 ， 2019
- 摘要：There have been many studies on the distribution of sensing tasks in mobile sensor networks. However, these studies rarely involve the problem that many agents in a mobile sensor network cooperate to perform complex sensing tasks. In order to address this challenge, first, we combined the structural characteristics of mobile sensor networks,the relationship between agents, and the relationship between agents and sensing tasks. Then we proposed the strength of cooperation between agents and the fitness of agents to sensing tasks, and discussed their roles in the dynamic distribution of sensing tasks in mobile sensor networks. Second, based on the above concepts, the two were unified as preference factors.In order to achieve the goal of improving task distribution efficiency, a task distribution algorithm based on random walk and cooperative relationship was proposed. At last, the comparison with the Personal Rank(PR) algorithm and HITS algorithm shows that the proposed algorithm has superiority in task distribution efficiency and accuracy.关键词：mobile sensor network;agent;sensing task;bipartite graph;task distribution116|328|2更新时间：2025-07-16
- 摘要：Aiming at the two basic problems, the error of interface interaction caused by hand gesture recognition errors and gesture non-recognition error, the intelligent teaching interface based on gesture interaction is designed and implemented.And the system can interact with the teacher by obtaining the gesture information of the teacher. The main innovation of this paper is to propose a flexible mapping interaction algorithm with multiple gestures mapping to the one semantic. The paper selects fourteen kinds of natural interactive gesture, and analyzes the common characteristics of various gestures mapping to the one semantic. The experimental results show that the proposed algorithm can effectively reduce the users' burden.The algorithm has been used in an intelligent teaching system interface based on gesture interaction.关键词：interactive mode;natural user interface;cognitive load;gesture interaction;human computer interaction108|266|1更新时间：2025-07-16
- 摘要：The system frequency accuracy based on a single atomic clock is difficult to break through 10-12 level, and the system based on multi-reference clock synthesis algorithm can't significantly improve the long-term stability and short-term stability of the clock at the same time, which is difficult to meet the requirements of ultra-high precision clock synchronization in many fields in the future. Combined with neural network algorithm, a local multi-reference clock synthesis algorithm based on back-propagation neural network (BP neural network) is proposed, which can improve the short-term and long-term stability of clock source at the same time.关键词：BP neural network;machine learning;multi reference clock synthesis;high precision clock145|303|2更新时间：2025-07-16
- 摘要：Design reuse is an important means to increase the productivity of analog and mixed-siganl IC designs. A gm/Id based resizing methodology for CMOS OpAmp's is proposed in this paper.The basic idea is to preserve the gm/Id parameters of some crucial transistors in the circuit with the aim of the approximate performance preservation of the resized circuit. The method to accurately match the gm/Id parameters based on the BSIM like model between the resized and the original circuit is presented. The resizing experiments of a two-stage Miller compensated OpAmp and a folded cascode OpAmp for the process migration from a 0.35 μm CMOS technology to a 0.18 μm ,0.13 μm, 90nm one have been performed to validate the proposed method. The simulation results show that the method generates the resized circuits with almost the same performance but reduced power and area consumption at a lower computational cost compared with the existing approaches.关键词：analog IC design;analog IP;design reuse;process migration128|227|0更新时间：2025-07-16
- 摘要：Compared with the conventional two-step (including parameter estimation and position solution) localization mode, direct position determination (DPD) algorithm has more advantages,such as higher position estimation accuracy, strong resolution capability, etc. However, the performance of DPD algorithm is affected by the array modeling errors. In this paper, the localization deviation for multiple signal classification (MUSIC)-based DPD algorithm is derived quantitatively by using first-order Taylor expansion in the presence of array modeling errors. The result demonstrates that there is a nonlinear relationship between the true location of the radiation source and biased position estimation obtained from MUSIC-DPD, but this relationship cannot be accurately expressed in practice. To solve this problem, we propose the use of a multilayer perceptron (MLP) neural network (NN) for localization deviation correction of DPD. This method can effectively correct the localization deviation of DPD caused by array modeling errors through adaptive learning of NN.关键词：passive location;array modeling errors;direct position determination (DPD);neural network (NN);first-order Taylor expansion148|231|3更新时间：2025-07-16
- 摘要：Communication in any form either verbal or non-verbal is vital to complete various daily routine tasks and plays a significant role in life. Facial expression is the most effective form of non-verbal communication and it provides a clue about emotional state, mindset and intention. Till now, facial expression recognition has been successfully applied to various fields such as safe driving, merchandise sales, clinical medicine, and so on. This thesis explores key techniques related to facial expression recognition. The main work and contributions are as follows. A dynamic facial expression recognition algorithm based on multi-visual descriptors and audio features is proposed under unrestricted conditions,in which dynamic facial feature extraction was conducted based on local spatial-temporal feature representation via multi-visual descriptors.Furthermore,the combination of video and audio features improves the recognition performance. Dynamic time warping based on timeline segmentation and covariance matrix proves to be effective in analyzing dynamic expression sequences of different time duration. To improve the generalization performance of facial expression recognition model, an integrated decision-making strategy based on weight voting by multiple individual recognition models is introduced. In order to effectively learning the weight for each individual recognition model, the method of voting weight learning by random re-sampling and the method of voting learning based on comparative advantages of individual recognition model are proposed. Finally the above ensemble model is applied and the recognition performance is further improved. Experiments on AFEW5.0 dataset validate the performance of the proposed dynamic facial expression algorithm.关键词：active expression recognition;multi-visual descriptors;ensemble model;weight learning76|204|3更新时间：2025-07-16
- 摘要：The studied technologies which were based on centralized physical devices and targeted on benefit optimization could not be applied in the enterprise network. To solve this problem, a cost optimization strategy based on location constrain (EL-VNE) was proposed for virtual network embedding in the enterprise network. Firstly, the computing and bandwidth capabilities were unified to a single node by using the complex theory, which decreased embedding times. Secondly, the EL-VNE model was presented and composed by mathematical descriptions about energy consumption, resource consumption and location constrain,which was proved to be NP-completed. Finally, heuristic algorithms were designed to find the solution of EL-VNE and then the embedding scheme with energy optimization was obtained. Compared with algorithms such as EAD and GLC ,EL-VNE saves energy consumption in the process of embedding and running, and has better performance.关键词：virtual network embedding;enterprise network;energy costs optimization;location constrain106|166|0更新时间：2025-07-16
- 摘要：Currently, action recognition methods can hardly carry out feature analysis,behavior classification,and action completion, and are incapable of accurately identifying the complete behavioral action sequence of intelligent agent for the discontinuous and incomplete motion capture, behavioral data missing or even broken in the time dimension, which are resulted from sensor device not being continuous coverage caused by the monitoring blind area. In this regard, we put forward a method of action recognition based on deep learning and artificial intelligence planning. Firstly, a deep learning network is constructed, by which the image is classified and trained using DRN(Deep Residual Network).After that,the extraction depth information of image frame feature for recurrent neural network is trained to enhance the classification effect. Secondly, the STRIPS (Stanford Research Institute Problem Solver) planning model is used to extract the image feature of deep learning, transforming into the description document for domain model, which facilitates deriving the optimal planning solution by means of forward state-space search planner. In the experiment, we exhibit that our method outperforms baselines in the public datasets, e.g., DCIGN(Deep Convolutional Inverse Graphics Networks), GAN(Generative Adversarial Networks), DBN(Deep Belief Networks), and SVM(Support Vector Machine).关键词：action recognition;deep learning;artificial intelligence planning;deep residual network;recurrent neural network;STRIPS planning model;forward state-space search planner157|357|6更新时间：2025-07-16
- 摘要：In a frequency division duplex (FDD) massive multiple input multiple output (MIMO) system, the compressed sensing (CS)-based channel state information (CSI) feedback methods cause a significant feedback overhead, due to the correct reconstruction of the support-set for the small magnitude elements of CSI. To reduce the feedback overhead,a CS-based CSI feedback scheme assisted by partial support-set is proposed in this paper, where the partial support-set of small magnitude elements and the compressed CSI are fed back to the base station (BS) together.Since the BS does not need to recover the support-set of small magnitude elements, the measurement requirement (i.e., the feedback overhead) to compress the CSI is greatly reduced. Compared with the conventional CS-based CSI feedback, the analysis and simulation results show that the proposed method can reduce CSI feedback overhead and computation complexity of CSI reconstruction, while guaranteeing the CSI recovery accuracy and achievable sum-rate.关键词：CSI;compressed sensing;massive multi-input multi-output;feedback overhead77|108|0更新时间：2025-07-16
- 摘要：The radio resources trading happened between the MTCGs and the LTE users. The credit degrees are built among the MTCGs according to the free radio resources trading based on the consortium blockchain. A consortium blockchain has been established on the BSs for publicly auditing and sharing of transaction records without depending on the trusted third parties. The resource transaction records are uploaded to the BS after encryption. After the transaction records pass the review and consensus process, the new blocks are stored on the BSs that can be accessed publicly by MTCGs, LTE users and the BSs connected to the consortium blockchain. In order to maximize the benefits of the system and support frequent resource transactions, a credit-based payment scheme and the corresponding optimal pricing strategy are proposed.关键词：consortium blockchain;credit loan;credit radio resources trading;credit degrees72|155|6更新时间：2025-07-16
- 摘要：Galileo system uses Interplex modulation to broadcast navigation signals at the E1 frequency. Due to the lack of prior information such as E1 signal power allocation and pseudo-code sequences, researchers can only carry out limited signal quality characteristics studies. Aiming at this problem, this paper proposes a signal-in-space quality assessment method based on high gain antenna, which realizes the analysis of E1 authorization signal and perfects the E1 signal evaluation system. The related power method is proposed to solve the signal component power ratio problem. The tracking result is used to solve the phase deviation estimation problem. A new time domain waveform separation method combining weighted combination averaging and code phase averaging is proposed, which overcomes the influence of message and code Doppler on time domain waveform characteristics evaluation. The S-curve offset Biases (SCB) parameters are used to quantitatively evaluate the signal ranging deviation. In this paper, the Galileo GSAT-0214 satellite is evaluated by this method. The results show that the signal component SCB of E1 is less than 0.2 ns, and the ranging performance is excellent. The multiplexing efficiency is 97.8%, which is better than GPS L1 signal and BOS-3 system B1 signal.关键词：signal-in-space;Galileo;authorized signal;power distribution;time waveform;correlation performance57|80|1更新时间：2025-07-16
- 摘要：Entity relation extraction is a crucial part of knowledge base construction. Among many methods of relationship extraction, the method of distant supervision combined with neural network model is satisfactory in terms of accuracy and other performance. However, there is often a large amount of noise data in the labeled corpus obtained by distant supervision, which has a great impact on the training of relationship extraction model. In this paper, we propose an entity relationship extraction model of convolutional neural network based on improved attention mechanism. Aiming at the sentence set containing the same entity pair, this model tries to find out all the positive instances that embody the relationship between the entity pair, construct the combined sentence vector, and discard the possible noise sentences, so as to minimize the impact of noise sentences and make full use of the semantic information of positive instances. Experimental results show that the accuracy of the proposed relation extraction model is better than that of the comparative relation extraction model.关键词：relation extraction;improved attention mechanism;convolutional neural networks;distant supervision;combined sentence feature vector344|224|12更新时间：2025-07-16
- 摘要：Aiming at the problem that the downlink channel estimation performance is limited due to the fast time-varying and non-stationary characteristics in the high-speed mobile environment, this paper proposes a channel estimation network based on deep learning,called ChanEstNet. ChanEstNet uses the convolutional neural network (CNN) to extract channel response feature vectors and recurrent neural network (RNN) for channel estimation.We use the standard high-speed channel data to conduct offline training for the learning network, fully excavate the channel information in the training sample, make it learn the characteristics of fast time-varying and non-stationary channels in high-speed mobile environments,and better track the characteristics of channel changing in high-speed environment. The simulation results show that in the high-speed mobile environment, compared with the traditional methods, the proposed channel estimation method has low computational complexity and significant performance improvement.关键词：OFDM;channel estimation;high-speed channel;deep learning;fast time-varying channel;non-stationary channel778|420|11更新时间：2025-07-16
- 摘要：The traditional one-class support vector machines are sensitive to noise data and not suitable for large-scale classification. In order to solve the problem, a novel one-class fuzzy support vector machine based on reduced convex hull called OC-FSVM-RCH is proposed for large-scale noise data classification. According to the reduced convex hull, OC-FSVM-RCH obtains the samples representing the geometric characteristics of normal class data in the kernel space. Then OC-FSVM-RCH improves the fuzzy support vector domain description algorithm, in which normal class data is enclosed in the smallest hypersphere,and the margin between abnormal class data and hypersphere is maximized. OC-FSVM-RCH can eliminate the noise at the edge of normal data contour and is insensitive to the noise inside the normal data. Experimental results show that the proposed algorithm achieves good results in terms of performance and training time.关键词：one-class;fuzzy support vector machine;reduced convex hull;noise data105|322|1更新时间：2025-07-16
- 摘要：In order to reduce the output distortion that caused by the non-constant group delay of digital filter, an equalization based optimization technique is proposed, in this paper, for decreasing the variation of group delay in cascaded infinite-impulse response digital filter. By inserting all-pass equalizer behind each stage of the cascaded ⅡR digital filter, the variation of group delay within the passband range can be reduced, and the output distortion of the digital filter can also be decreased. For the group delay optimization method proposed in this paper, when the 1st and 2nd order equalizers are used for circuit optimization, the variation of group delay is reduced by 28.19% and 49.93% respectively in the passband range of 0~100Hz. Based on the 0.18μm CMOS standard cell library for logic synthesis and layout design, the area of the entire filter circuit IP core layout is 0.1747mm2. Compared with the existing literature methods, this method has a significant effect on group delay optimization, and the power consumption and area are small in circuit implementation, which is very suitable for system-on-chip applications.关键词：digital filter;IIR(Infinite Impulse Response);cascaded;group delay;all-pass equalizer156|244|1更新时间：2025-07-16
- 摘要：IP address dynamic protection techniques will introduce additional overhead. Therefore, the performance of normal network transmission decreases. A dynamic protection system of IP address accelerated by Vector Packet Processing (VPP) is designed and implemented for the first time, which can hide the real IP address and enhance the system's data Processing ability. Firstly ,fast forwarding logic and slow forwarding logic are designed respectively for different logic of control plane and data plane processing, so as to minimize the number of copies in data message processing. Secondly, facing the frequent mapping between real IP and virtual IP, an efficient dynamic IP address transformation mechanism of Shared memory is proposed. Thirdly, the optimization algorithm is used to formulate the IP hopping strategy, and the hashing chain algorithm is used to formulate the efficient virtual IP address pre-allocation mechanism. Minimize system performance losses. Finally, the experimental results show that the system can effectively resist DoS attacks and control the potential detection attack hit rate below 16%, which is significantly improved in data processing performance.关键词：network active defense;IP mutation;vector packet processing;optimization64|176|1更新时间：2025-07-16
- 摘要：Since computational complexities of the existing methods such as classic GN algorithm are too costly to cluster large-scale graphs, this paper studies sampling algorithms of large-scale graphs, and proposes a clustering-structure representative sampling (CRS) which can effectively maintain the clustering structure of original graphs. It can produce high quality clustering-representative nodes in samples and expand according to the corresponding expansion criteria. Then, we propose a fast population clustering inference (PCI) method on the original graphs and deduce clustering assignments of the population using the clustering labels of the sampled subgraph. Experiment results show that in comparison with state-of-the-art methods, the proposed algorithm achieves better efficiency as well as clustering accuracy on large-scale graphs.关键词：large-scale graphs;graph sampling;graph clustering;population inference;clustering representative nodes;expansion criteria301|239|2更新时间：2025-07-16
- 摘要：To improve the accuracy of segmenting the tumor sub-regions in glioblastoma multiforme (GBM) multi-modal magnetic resonance (MR) images, a GBM multi-modal MR images automatic segmentation algorithm is proposed by using multi-class convolution neural network (CNN). Firstly, after 98% winsorization and registration for the GBM multi-modal MR images, the bias field was corrected by using the N4ITK method. Secondly, a multi-class CNN model mainly consisting of four convolutional layers, two pooling layers and two fully connected layers was constructed; the GBM multi-modal MR images were pre-segmented after training,and voxels were classified into five different labels.Finally,all false positive regions smaller than 200 voxels were removed,and the final segmentation results were obtained by median filtering. The Dice similarity coefficient DSC,positive predictive value PPV and average Hausdorff distance AHD were adopted as the evaluation index, and the DSC, PPV as well as AHD were 0.889±0.087, 0.859±0.127 and 1.923 for segmenting the entire tumor tissues in F-C-GBM dataset by the proposed algorithm, respectively. Results indicate that the proposed method can effectively improve the performance in the segmentation of the GBM multi-modal MR images and may be expected to have clinical application prospects.关键词：glioblastoma multiforme;multi-modal magnetic resonance image;automatic segmentation;multi-class convolutional neural network;image patch59|200|7更新时间：2025-07-16
- 摘要：With the rapid development of the Internet of things and the body area network, the human motion tracking technology has been widely used in medical, security and many other fields. A human motion tracking method based on IMU/TOA fusion is proposed to solve the problem of error accumulation and drift in a single IMU human motion tracking system. On this basis, the effectiveness of IMU/TOA fusion method is proved theoretically by deducing the cramer-rao lower bound (CRLB) of fusion system. The experimental results show that the proposed human motion tracking method based on IMU/TOA fusion has great improvement in both spatial and temporal performance.关键词：body area network (BAN);Internet of Things (IoTs);human motion tracking (HMT);inertial measurement unit (IMU);time of arrival (TOA);Cramer-Rao lower bound (CRLB);Information fusion264|153|1更新时间：2025-07-16
- 摘要：The fully dielectric isolation structure of Silicon-on-Insulator (SOI) devices improves their immunity to single event effects, but also makes them more sensitive to total ionizing dose (TID) effects. In order to evaluate the sensitivity of SOI devices to TID effects, a simulation method based on TCAD (Technology Computer Aided Design) was proposed. By modeling the three-dimensional structure of SOI devices and using the TCAD built-in radiation model to simulate the generation, transport and capture of radiation-induced charges in oxide layer, the effects of radiation-induced trap charges in buried oxide (BOX) and shallow trench isolation (STI) oxide on the electrical properties of SOI devices are evaluated respectively. Based on this simulation technique, the effects of radiation bias,channel length, body doping concentration and STI morphology on TID effect of SOI MOSFET were investigated. The simulation results show that higher body doping concentration, smaller STI divot depth and steeper STI sidewall will be helpful to improve the TID hardness performance of SOI devices.关键词：silicon on insulator;total ionizing dose effect;shallow trench isolation;technology computer aided design117|258|1更新时间：2025-07-16
- 摘要：High-altitude electromagnetic pulse may cause damage to electronic equipments through coupling pipeline system. In order to study the effects of high-altitude electromagnetic pulse on overhead pipeline, the electromagnetic field-to-overhead pipeline coupling equations are derived from Maxwell's equations. The coupling equations are solved in frequency domain by applying the Green's functions, and the time-domain waveforms of induced currents and voltages are obtained by using inverse Fourier transform. Finally, the currents and voltages on the pipeline induced by intermediate-time HEMP(E2) based on IEC standard waveform are calculated.关键词：high-altitude electromagnetic pulse;overhead pipeline;induced current;induced voltage46|109|3更新时间：2025-07-16
- 摘要：A teaching-learning-based optimization algorithm with rectangle neighborhood structure (RNTLBO) is proposed to overcome the shortcomings of low global search precision and premature convergence of the original teaching-learning-based optimization algorithm (TLBO) while handling complex multimodal functions. In the algorithm, the population space is designed as a rectangular structure, and the individual rectangular neighborhood is determined by the rectangle thickness and the individual rectangular region surrounding it. In both teaching and learning stages, the optimal individual in the neighborhood is used to guide the search, which strengthens the ability of the algorithm to explore new solutions and exploit local optimal solutions.In order to prevent the algorithm from falling into the local optimum prematurely, the individual perturbation stage guided by search boundary information is added, so that the population can maintain good diversity even in the later evolution stage. The simulation results of complex functions with shift and rotation show that the new algorithm is superior to the original TLBO and some other recently improved variants in terms of accuracy and stability in most cases.关键词：teaching-learning-based optimization (TLBO);rectangle neighborhood structure;layer number of neighborhood;boundary information;individual disturbance;population diversity97|148|3更新时间：2025-07-16
- 摘要：Fuzzy c-means (FCM) algorithm is a clustering process based on soft partitioning,and it has been widely used in machine learning, image processing and computer vision. Although a large number of image segmentation algorithms based on FCM have been proposed, it is still a challenge research topic to research image segmentation based on fuzzy clustering. In this paper, image segmentation algorithms based on FCM are roughly grouped into three categories: FCM algorithms based on spatial neighboring information, FCM algorithms based on histogram information,and FCM algorithms based on dimension weight. We firstly analyze and elaborate the current research on FCM algorithms. Afterwards, we analyze the performance of different algorithms according to experiments.Finally,we conclude the drawbacks of image segmentation algorithms based on FCM and the future research direction.关键词：image segmentation;fuzzy c-means clustering;neighboring information;histogram;hidden markov model254|515|13更新时间：2025-07-16
- 摘要：Wireless energy harvesting is regarded as a promising energy supply alternative for energy-constrained wireless networks. In this paper, we investigate secure transmission in RF-powered energy harvesting relay networks, where an energy-constrained relay is untrusted. Considering that the energy-constrained relay can harvest energy from RF signals to forward the confidential messages, the source node with multiple antenna employs transmit antenna selection (TAS) to enhance the energy harvesting efficiency of the relay, and the destination node employs artificial interference to enhance security. Closed-form expressions of the secrecy outage probability, connection outage probability and secrecy throughput have been derived. Numerical results verify the accuracy of analytical expressions, and reveal the influence of the system parameters on the secrecy performance.关键词：RF-powered energy harvesting;untrusted relay;physical layer security;secrecy throughput42|121|2更新时间：2025-07-16
-
FPGA Implementation of Real-Time Detection for Weak Electromagnetic Pulse Signal with Low Distortion
摘要：In order to detect weak electromagnetic pulse signals in wideband noise at a real-time rate, the detection algorithm and FPGA implementation were researched. According to the phenomenon of general "in-well out-of-synchronization" in the strongly coupled Duffing oscillators caused by the pulse signal, the original weak pulse signal could be detected through the principle of variable difference. The Duffing differential equations were solved by a real-time algorithm named fourth-order Runge-Kutta, which has the advantage of maximum stability domain and minimum truncation error. We devised a pulse detection pipeline structure that was suitable for running on the FPGA. In order to improve the detection rate,a data parallel structure was designed to calculate 73 segments of data simultaneously. To avoid the problem of waveform distortion caused by parallel computing, we separated the threshold detection process from the waveform extraction process by adding a FIFO memory. Experimental results show that the minimum detectable signal-to-noise ratio can reach-18dB in the detection of weak lightning electromagnetic pulses.关键词：coupled Duffing oscillators;weak pulse signal;real-time detection;FPGA(Field Programmable Gate Array)66|190|1更新时间：2025-07-16 - 摘要：Based on the 0.25μm GaN HEMT process platform of Nanjing Institute of Electronics Technology, an ultra-wideband power amplifier with an operating frequency of 1~4GHz and a continuous wave output power greater than 80W is designed. The amplifier uses a low-pass LC matching network to achieve the conversion of the input and output impedance of the die to the real impedance, and the ultra-wideband matching is realized by the Chebyshev converter structure; after the single input/output port is matched to 100Ω, the two direct circuits are synthesized to 50Ω. The method achieves power synthesis of a high-power ultra-wideband power amplifier. The amplifier has a bias voltage of 32V and a quiescent current of 0.4A. The test results show that in the bandwidth of 1~4GHz, the continuous wave output power of the amplifier is greater than 49.05dBm (80.3W), the maximum output power is 50.6dBm (114.8W), the saturation power gain is greater than 9dB, and the power flatness is less than ±0.8dB.The maximum drain efficiency is 62.5%.关键词：ultra wideband;chebyshev converter structure;GaN;internal matching140|249|1更新时间：2025-07-16
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09064830号-19
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0