最新刊期
卷 49 ， 期 1 ， 2021
- 摘要：Online traffic classification is getting into troubles when the network applications are exploding. The traditional machine learning methods based on statistical characteristics of packets work well in stable network environment, but not in congestion environment with serious delay and packet loss. Therefore, a novel classification method based on granular computing is proposed in this paper. Granular computing belongs to the field of artificial intelligence computing, which is usually used to process missing, incomplete or noisy data. So we first define granules for the traffic, then construct the relations between the granules, and finally establish the relation matrix. The traditional statistical characteristics are only the special case of the relation matrix when the scale is the largest. The granular relation matrix can describe the traffic more comprehensively and classify them more accurately. The experiment results show its validity and advantages when compared with other methods.关键词：network traffic;online classification;granular computing;relation matrix;difference degree53|138|0更新时间：2025-12-08
- 摘要：Toward the co-channel signal interference problem in wireless communications, inspired by sparse representation and cyclostationary characteristics, the dictionary matrix is constructed by establishing the sample set of position coordinates, and the sparse signal is reconstructed in cyclic frequency domain using optimization method. Then, the position of the transmitter is directly determined from the cyclic spectrum slice of the received signal. Simulation results show that the localization performance of the proposed algorithm is close to the derived Cramer-Rao lower bound, and outperforms the existing frequency domain direct localization algorithm and two-step localization algorithm, especially in cases of low signal-to-noise ratios and signal-to-interference ratios.关键词：transmitter localization;co-channel interference;cyclic spectrum;sparse representation;sparse signal reconstruction52|81|0更新时间：2025-12-08
- 摘要：In order to improve the sound field synthesis quality of the sound source in the region of interest, a non-uniform secondary source distribution method for sound field analysis and synthesis is proposed. First, the sampling of the continuous secondary source is divided into two different sampling regions. With a certain number of loudspeakers, the number of samples in the region of interest is increased and the number of samples outside the region of interest is reduced. Then, using the Hermitian characteristic of the sampling aliasing matrix, Rayleigh entropy is used to estimate the maximum error of different non-uniform sampling schemes to guide the selection of non-uniform sampling schemes. Finally, the frequency domain calculation method of the driving signal for the non-uniform secondary source distribution is obtained. The experimental results show that the proposed non-uniform distribution method improves the aliasing-free radius by 12.9%~22.6% for the sound sources from the region of interest under the 2.5D circular boundary condition, compared with the optimal modal bandwidth of the uniform distribution. The effectiveness of the proposed method is verified.关键词：acoustic signal processing;non-uniform sampling;sound field synthesis;spherical harmonics77|61|0更新时间：2025-12-08
- 摘要：According to the characteristics of high dynamic aircraft platform terminal, this paper studies the problem of frequency accurate estimation in high dynamic aircraft platform terminal communication, and proposes a decomposition algorithm for frequency estimation accurately. Firstly, the approximate position of pilot frame is determined and approximate value of frequency offset is estimated based on the segmented search. Secondly, the accurate position of each date frame is determined by the frequency offset search method, and the higher frequency offset estimation accuracy is obtained by which data frame is de-expanded. Finally, the PN field in data frame is used to further improve frequency offset estimation accurately, and frequency offset equation and historical data are combined to form a set of equations, by which the change rule of frequency offset is obtained and tracked. It is verified that the algorithm proposed in this paper can accurately estimate the frequency offset in terminal communication of high dynamic aircraft platform, which can meet requirement of time-frequency synchronization.关键词：high dynamic aircraft;terminal;accurate estimation;frequency offset33|219|1更新时间：2025-12-08
- 摘要：The auto-regressive (AR) model is an effective method to describe the correlation of time series. The classic AR coefficient estimation method utilizes a simple assumption about residual signal. It is a challenge to accurately estimate the auto-regressive coefficients in a complex environment such as noise or interference. Even though Deep Neural Networks (DNN) based AR (DNN-AR) coefficient estimation method can estimate the AR coefficients in a complex environment, the DNN-AR method is easily affected by the numerical stability of Levinson-Durbin recursion (LDR) approach during the training stage. The main target is to improve the stability and overall performance of the DNN-AR based method. In this paper, the precision transform method is utilized to improve computational efficiency while keeping system stability, and the generalized analysis-by-synthesis combing DNN (GABS-DNN) model is proposed for improving the accuracy of AR coefficient estimation and stability of the DNN training in the noisy environment. The GABS-DNN model consists of three main parts: Spectrum enhancement network in the modifier, DNN preprocessing and LDR parameter estimation at the encoder, and the conversion from autoregressive coefficient to power spectrum at the decoder. In the process of optimizing the objective function, the error between the enhanced spectrum and the observed spectrum is added for reducing the influence of the gradient of the LDR on the enhanced network during back-propagation, which results in a stable estimation of the AR coefficients of noisy speech.关键词：auto-regressive (AR) coefficients;generalized analysis-by-synthesis (GABS);deep neural networks (DNN);Levinson-Durbin recursion (LDR)34|151|0更新时间：2025-12-08
- 摘要：In this paper, we propose a blind NSST domain image watermark decoder, wherein a vector-based HMT statistical model using BKF distribution is used. In the proposed scheme, the NSST is firstly performed on the original host image, and then the adaptive high-order watermark embedding strength functions are constructed, and finally the watermark data is embedded into the significant high-frequency coefficients in NSST domain. At the watermark receiver, NSST highpass coefficients are firstly modeled by employing the BKF vector HMT, where the BKF marginal statistics and strong intra-subband, cross-scale, and cross-orientation dependencies of NSST coefficients are incorporated. Then the statistical model parameters of BKF vector HMT are estimated using the expectation maximization approach. And finally a blind image watermark decoder is developed using BKF vector HMT and the maximum likelihood decision rule. The experimental results validate the effectiveness of the proposed technique.关键词：image watermarking;vector-based HMT;BKF distribution;nonsubsampled Shearlet transform;adaptive watermark embedding strength;maximum likelihood decision37|51|0更新时间：2025-12-08
- 摘要：Secure multi-party computation (SMC) is an important research direction of cryptography. In this paper, we study the secure computation of intervals. Using the Paillier encryption scheme, we design the protocols of relationship between an interval and a point (or an interval). Firstly, the outputs of protocols are ciphertexts. If we extend it to rational intervals, the protocols are safer and more efficient than existing protocols. And then, we study the multi-dimensional problems, that is, the threshold problems of multiple points (or intervals) and intervals, which are new problems in SMC. Since the outputs of the basic protocols are ciphertexts, the multi-dimensional problem protocols are more secure. We strictly prove the security of the protocols using the simulation paradigm method, analyze and demonstrate the efficiency of the protocols through experiments, and compare with the related work to illustrate that the protocols are efficient.关键词：cryptography;secure two-party computation;relationship between point and interval;relationship between interval and interval;threshold problem25|40|1更新时间：2025-12-08
- 摘要：In the electromagnetic compatibility analysis and signal integrity analysis, it is difficult to do full-wave electromagnetic simulation of system-in-package (SiP) due to its complex, multi-scale and multi-material characteristics. Here, we present a high-performance program JEMS-CDS developed on parallel infrastructure JAUMIN and its high-efficiency algorithms in SiP electromagnetic simulations. A positive definite matrix is constructed by analyzing the FEM matrix built for SiP simulation. Numerical results show that better convergence performance than traditional ones can be achieved, as well as a high parallel efficiency when scaling to hundreds of CPU cores with nearly one hundred millions unknowns. Finally, a practical product is simulated to acquire the high-resolution distribution of electromagnetic, supporting the analysis of electromagnetic compatibility and signal integrity.关键词：finite element method;preconditioned method;parallel computing;system-in-package;cross talk31|41|3更新时间：2025-12-08
- 摘要：Recently, 3D model retrieval based on views has become a research hotspot. In this method, 3D models are represented as a collection of 2D views, which allows deep learning techniques to be used for 3D model classification and retrieval. However, current methods need improvements on both accuracy and efficiency. We propose a 3D model retrieval method, which includes index building and model retrieval. In the index building stage, representative views are selected and input into a well-learned Convolutional Neural Network (CNN) for feature extraction and classification. Next, the features are organized according to their labels to build indexes. In the retrieval stage, the representative views of the input model are classified into a category with the CNN and voting algorithm, and then only the features of one category rather than all categories are chosen to perform similarity matching. In this way, the searching space for retrieval is reduced. In addition, the number of the used views for retrieval is gradually increased. Once there is enough evidence to determine a 3D model, the retrieval process will be terminated ahead of time. Experiments on the rigid 3D model datasets ModelNet10, ModelNet40, and the non-rigid 3D model dataset McGill10 show that the proposed method can improve the retrieval efficiency substantially while keeping high retrieval accuracy rates at 94%, 92% and 100%, respectively.关键词：model retrieval;model classification;representative views;convolutional neural network11|54|4更新时间：2025-12-08
- 摘要：An improved low-light image enhancement (LLIE) algorithm based on the hybrid strategy of deep learning and image fusion was proposed in this paper. We first adopted illumination prediction model to quickly estimate the optimal illumination component from a given low-light image and obtain its corresponding moderately exposed image within the framework of the Retinex model. Then the low-light image and its over-exposed image were used as supplementary images for the moderately exposed image. Finally, the three images were fused within the framework of the local structured fusion and the chrominance weighted fusion mechanism to obtain the final enhanced image. Experimental results demonstrate that, compared with the state-of-the-art LLIE algorithms, the proposed hybrid strategy has significant advantages in both subjective and objective image quality evaluation metrics with better image edge preservation and color fidelity effect on local image details.关键词：low-light image enhancement;deep learning;illumination predict model;moderately exposed image;over-exposed image;image fusion21|84|5更新时间：2025-12-08
- 摘要：Coverage is an important metric for evaluating the sensing quality of visual sensor networks (VSNs). Unlike traditional coverage issues, full-view coverage requires to capture target’s effective face from any direction, and its coverage estimation issue is more complicated. At present, most of literatures assume that a great number of homogeneous visual sensors are randomly scattered in the monitoring area to achieve full-view. This paper studies the full-view estimation problem in heterogeneous VSNs. In order to eliminate the boundary effect, the concepts of extended monitoring area and maximum detection area are introduced. To evaluate the performance of the proposed full-view estimation expression, several simulation experiments are conducted to validate the theoretical results. The results show that the mean absolute coverage error between the theoretical values and the experimental values is controlled less than 6.5%.关键词：heterogeneous visual sensor network;coverage estimation;full view coverage;node prediction9|60|3更新时间：2025-12-08
- 摘要：Based on the hierarchical model of SDI (Software-Defined Intelligence), a system for action recognition during sleep is designed to deal with various changing factors in smart environment through rule-based reasoning. A time queue is designed to extract the characteristics of actions in real-time to train the model, and a rule extraction algorithm is proposed to extract the rules required by the system from the model. Depending on these rules,the proposed system can recognize nine types of sleep actions: The recognition precision of each type can exceed 96%; the total recognition accuracy can reach 98.9%. Importantly, it has more robust adaptability than other systems. Experimental results show that the system can update rules for quickly adapting to changes in node position, the number of nodes, and user requirements.关键词：smart environment;software-defined intelligence;sleep action recognition;rule-based reasoning8|46|1更新时间：2025-12-08
- 摘要：From the perspective of geometry, based on the geometric interpretation of the four-point binary interpolating subdivision scheme, this paper analyzes the geometric meaning of the four-point ternary interpolating subdivision scheme, and modify the scheme to combine approximating subdivision; then a blending ternary subdivision scheme with parameters is obtained. Many existing interpolating subdivision schemes and approximating subdivision schemes can be seen as special cases of this scheme. We also use generating polynomial method to analyze the Ck continuity of limit curve produced by the blending subdivision. A new C4 continuous five-point ternary curve subdivision scheme is obtained. Numerical examples show that the proposed blending subdivision scheme can be used to control the shape of limit curves by selecting appropriate parameters.关键词：ternary subdivision scheme;blending;interpolating subdivision;approximating subdivision;Ck continuity14|18|0更新时间：2025-12-08
- 摘要：With the rapid development of mobile Internet technology, the traditional recommender system has not been well adapted to the location-based recommendation service, and it also faces the risk of privacy leaks. In this paper, a distributed privacy preserving recommendation framework is proposed, and a singular value decomposition recommendation algorithm based on distributed framework is designed by using the differential privacy theory. Furthermore, we use order preserving encryption function to protect user request location. Theoretical analysis and experiments on two real datasets show that the proposed method not only has stronger privacy protection ability, but also has better recommendation performance than traditional recommendation algorithms.关键词：recommender system;distributed framework;location-based service;privacy preserving;order preserving encryption32|39|3更新时间：2025-12-08
- 摘要：Improving micro-electro-mechanical inertial navigation system based indoor pedestrian localization is a hot and hard topic these days. It is because random errors of MEMS gyroscopes and accelerators are very complicated, and very difficult to model using norm techniques. In this paper, a map matching method based on conditional random fields and indoor 3D map is proposed. The algorithm is designed by establishing the mathematical relationship between the inertial position and the interior points of indoor map, and the optimal trajectory algorithm is used to get the estimated locations. The position errors of inertial navigation are corrected according to the matching points solved by the algorithm.关键词：micro-electro-mechanical inertial navigation system(MEMS-INS);indoor pedestrian localization;conditional random fields;indoor map matching7|54|3更新时间：2025-12-08
- 摘要：With the expansion of the application of blockchain technology in the socio-economic fields, increasing attention has been paid to the security of blockchain. This paper proposes a security risk assessment method for blockchain, which quantifies the security risk of blockchain from the aspects of technical architecture and hash rate. At first, we build the Blockchain Trusted Computing Base (BTCB) model based on the technology architecture. Then, we design a sensitivity analysis method combined the AHP and paired comparison to assign security weights to each security risk factor. Finally, we construct a security risk assessment model. In the experimental part, we use this model to score 15 common public-chain blockchain projects, and conducte comparative analysis with the evaluation data of four blockchain rating agencies with high market recognition. The experiment results verify the feasibility of this method.关键词：blockchain;risk assessment;BTCB(blockchain trusted computing base);AHP (analytic hierarchy process);hash rate70|51|6更新时间：2025-12-08
- 摘要：Since the result of classic evidence theory in dealing with high conflict situation is unreasonable, after analyzing the reasons and various improved methods, the evidence combination method based on conflict relation network is proposed. The method considers direct and indirect relations adequately, takes advantage of the idea of PageRank algorithm to obtain the weight of evidence by building network. Then the evidence is revised and weighted. Finally, the fusion is accomplished by adopting the combination rule of Dempster. It solves the problem above preferably while keeping the good mathematical properties of evidence theory. Analysis of examples indicates that, the proposed method can objectively reflect uncertain information with conflicting evidences well.关键词：evidence theory;conflict relation network;weight39|39|3更新时间：2025-12-08
- 摘要：Graph aggregation is a technology that aggregates a large-scale graph into a compact and small-scale graph while retaining the structure and attribute information of the original graph. With the increasing size of graph, graph data becomes difficult to query and store. Distance-based queries, such as shortest path queries, depend heavily on the size of graph.In this paper, a distance query-oriented attribute weighted graph aggregation algorithm is proposed, which not only guarantees the similarity of structure and attributes between nodes, but also preserves the distance between nodes, and effectively reduces the size of the graph. The experiments prove that this method is effective and efficient in the query tasks.关键词：graph aggregation;graph query;distance preserving;structure similarity;attribute entropy21|22|0更新时间：2025-12-08
- 摘要：For the mainstream lip motion and voice coherence judgment method, the whole sentence (segment) is analyzed without screening the content. This leads to large dictionary size and high computational complexity, and the result is vulnerable to weak related segments such as mute. Considering the vowel with significant lip shape changes as a representative pronunciation event and combining with the statistical results of the audio-visual initial delay distribution range, a consistent decision method based on audio-visual matching of vowel pronunciation events and position delay analysis is proposed. Firstly, the dictionary learning data is selected by the proposed audio-visual vowel segmentation method, and then the vowel dictionary is used to analyze the matching of the vowel event, and the time delay distribution of each vowel position is statistically scored. A consistency judgment is made by a scoring mechanism in which the vowel pronunciation event lip matching score and the position delay analysis score are combined. Experimental results show that the proposed method is superior to compared algorithms in recognition performance and reduces the amount of computation compared with the traditional dictionary method.关键词：coherence analysis;initial/final segmentation;dictionary learning22|38|0更新时间：2025-12-08
- 摘要：Container technology improves the efficiency of application distribution and deployment with its features of lightness, flexibility and rapid deployment. However, the characteristics of low resource isolation and shared kernel introduce new security risks to containers and cloud platforms. This paper proposes an anomaly detection scheme of processes behavior in container based on system call sequences and long short-term memory (LSTM) neural network, the scheme collects the system call sequence data of the whole life cycle of processes through the agentless monitoring mode, and uses LSTM to capture the semantic features of sequences. At the same time, two methods of abnormal decision are proposed by means of cumulative deviation in local window. Furthermore, in order to optimize the training efficiency of the model, an algorithm for removing duplicate short sequence samples with the same ratio is designed. The experimental results on the public dataset and real attack scenarios show that the scheme can effectively detect the abnormal behavior of processes in container, and the detection performance is better than other similar methods.关键词：anomaly detection;container;long short-term memory;system call;neural network41|58|5更新时间：2025-12-08
- 摘要：Based on the computation and storage resources limitation of single edge node and the demand for efficient computing services in big data scenarios, this paper proposes a deep reinforcement learning based cloud-edge collaborative computation offloading mechanism. Specifically, based on a comprehensive consideration of computing resources, bandwidth and offloading policy, an optimization problem is formulated to minimize the weight sum of execution delay and energy consumption of all user tasks. An asynchronous cloud-edge collaborative deep reinforcement learning (ACEC-DRL) algorithm is proposed to solve such optimization problem. This algorithm can effectively satisfy the demand of efficient computing services in big data scenario by jointly leveraging the computation capabilities of cloud and edge nodes. Meanwhile, under the various and dynamic environments of edge nodes in the edge cloud, this algorithm can adaptively adjust offloading policy to achieve the minimization of system cost. Finally, the extensive simulation results show that the proposed ACEC-DRL algorithm has the characteristics of fast convergence rate and high robustness, and its optimal offloading policy closely approximates to the solution of greedy algorithm with the lowest computation cost.关键词：deep reinforcement learning;edge computing;computation offloading;resource allocation;energy consumption48|53|8更新时间：2025-12-08
- 摘要：In order to optimize the efficiency of three-phase inverter, a three-phase resonant inverter with low-loss auxiliary circuit is proposed, in which auxiliary circuit is set up on the bridge arm of each phase. When the auxiliary resonant circuit works in the commutation process of the inverter, the main switches can achieve zero-voltage soft-switching, and the auxiliary switches can achieve zero-current soft-switching. The switching loss is obviously reduced. In addition, after the resonance state of the auxiliary circuit ends, the remaining electric energy in the resonant inductance will be fed back to the energy-storage capacitance through only one diode, reducing the loss of the auxiliary circuit when the electric energy is fed back, which is beneficial to the realization of the low-loss of the auxiliary circuit. In this paper, the working flow of the circuit is illustrated. The experimental waveforms show that both the main switch and the auxiliary switch achieve soft-switching. The research findings have reference for the research and development of high-efficiency three-phase inverter.关键词：inverter;bridge arm;auxiliary circuit;resonance;low-loss;soft-switching;switching loss45|45|1更新时间：2025-12-08
- 摘要：A broadband tunable metamaterial absorber based on vanadium dioxide (VO2) rings with different radius loaded on the dielectric layer is designed. The phase transition characteristics of VO2 with temperature are used to realize the dynamic adjustment of the absorption curve of external temperature. Simulation calculations show that the absorber can reach more than 90% in the bandwidth of 8.09~11.23THz when the external temperature is 350K, showing high absorption characteristics; while the absorption coefficient in the same frequency band is always lower than 20% when the external temperature is 300K, achieving the adjustable function of electromagnetic wave absorption. The equivalent impedance and electric field distribution of the absorber are further analyzed and discussed, and the adjustment mechanism of VO2 on the absorption is clarified. In addition, the article discusses the effects of structural parameters, polarization angles, and angles of incidence on the absorptivity. The results show that reasonable selection of parameters of the structure can achieve the independence of the absorbing performance from the polarization angle and the angle of incidence. The conclusions of this paper have promising potential for designing other types of ultra-wideband tunable absorbers.关键词：metamaterial absorber;terahertz;broadband;tunable11|28|3更新时间：2025-12-08
- 摘要：An algorithm for estimation of direction of arrival (DOA) and range of near field source based on the spatial differential technique is proposed in this paper. The algorithm firstly utilizes the feature that the stationary noise covariance matrix is symmetrical about the main diagonal and constructs the spatial difference matrix only containing the target signal location information. Then, it proves the distribution characteristics of the matrix eigenvalues and selects the noise subspace reasonably. Finally, the DOA and range estimations for near-field sources can be obtained through the spectral searching. The algorithm can effectively suppresses the unknown stationary noise and avoid the pseudo peak problems for the application of the spatial differential method when used to solve the source localization. Computer simulations confirm the satisfactory performance of the proposed algorithm.关键词：source localization;spatial differential technique;stationary noise;multiple signal classification(MUSIC)9|29|2更新时间：2025-12-08
- 摘要：Video object detection and tracking algorithms have become the research focus in the field of computer vision. Traditional methods need to manually collect samples to train detection models, and build object detection and tracking systems. When the imaging conditions change, it is necessary to re-collect samples to train the detection model and re-adjust the entire detection and tracking system, which requires tedious human efforts. In this paper, a multi-object detection and tracking algorithm is proposed based on few-shot learning. With this approach, a hybrid classifier that models one object class can be generated by simply marking several bounding boxes around the object in the first video frame. An online gradual learning algorithm is proposed to learn the object pose changes and update the model. Combined with the color-based object tracking algorithm, our method automatically builds high-precision object detection and tracking systems without manual collection and labeling training samples. This approach can be conveniently replicated many times in different surveillance scenes and produce scene-specific detectors under various camera viewpoints. Experimental results on several video datasets show our approach achieves comparable performance to robust supervised methods, and outperforms the state-of-the-art online learning methods in varying imaging conditions.关键词：few-shot learning;multi-object detection;multi-object tracking;online learning23|81|2更新时间：2025-12-08
- 摘要：Blockchain is a distributed ledger that integrates key technologies such as distributed storage, peer-to-peer transmission, consensus mechanism, cryptographic algorithm and smart contract. It has the characteristics of decentralization, non-tampering, transparency and so on. In recent years, security problems of blockchain have gradually emerged, hindering the development of blockchain applications. This paper introduces the basic concept and security model of the blockchain, and analyzes the security problems of the blockchain. Then, based on the attribute-based cryptography, the various researches of the blockchain security technology are analyzed from the three aspects of access control, key management, and data privacy protection, and the characteristics of the main solutions are discussed. Finally, we summarize the research progress of the blockchain security technology based on the attribute-based cryptography, and discuss the future research work.关键词：blockchain security;cryptography;attribute-based encryption;attribute-based signature15|59|8更新时间：2025-12-08
- 摘要：This paper proposes software architecture reconstruction technology based on layered architecture pattern recognition. The input of the recognition is the source code and the unnecessary source code will be filtered out at first. Then the approach relies on lexical information from the source code to mine the semantic relation between system entities and using a topic model to extract the responsibility of entities, which is then used to cluster these entities into cohesion components. Later, the approach supplements the structural information between components to generate the component graph and use the ILD property to recognize the actual software layers. Based on the results of pattern recognition and the principle of layered pattern, position the nonstandard existing in the system as the reconstruction point, and relevant reconstruction suggestions to assist the designers and developers in the reconstruction implementation. Finally, this paper selects 10 open source software systems in Github and SourceForge as experimental objects to verify the effectiveness of the technology in this paper. This technology can greatly improve the effectiveness of pattern recognition, the recognition effect of illegal refactoring points and the implementation effect of refactoring suggestions. This technology can also assist developers in the implementation of architecture reconstruction to a certain extent, improve the situation of the system violations, and improve the quality of the software.关键词：architecture refactoring;layered pattern;recognition of architectural patterns;refactoring point positioning39|53|0更新时间：2025-12-08
- 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