最新刊期
卷 47 ， 期 5 ， 2019
- 摘要：Due to deficiencies stemmed from its original design,the current Internet has exposed many serious drawbacks in service quality,resource utilization,and management/control capabilities,which cannot satisfy increasingly complex and diversified requirements from "Internet +" businesses on communication networks and services.Therefore,how to design a clean-slate Internet has become one of the most urgent and core research contents in the information field,and it has received great attention from both academics and industries.Hence,in this paper,we propose a novel network architecture with its associated key mechanisms,namely Smart Integration Identifier Networking (SINET-I),based on our previous works of Identifier Networking and Smart Identifier Networking.By smart integration of network-wide multi-space and multi-dimensional resources,SINET-I is able to achieve on-demand provision of individualized services and effective support for flexible networking,thus,it can provide differentiated and customized communication network services efficiently for various industries and users,thereby applicable to many different application scenarios such as high-speed rail networks,industrial Internet and vehicular networks.关键词：identifier networks;smart integration networks;future Internet258|499|6更新时间：2025-07-16
- 摘要：Existing incremental fuzzy clustering algorithms which can be used for time series often require setting multiple control parameters.To solve this problem,a fuzzy clustering algorithm of time series based on adaptive incremental learning is proposed.First,the cluster structure information obtained by the previous clustering process is inherited to initialize the current clustering process.Then,the outliers in current data block are adaptively searched without parameters,and new clusters are automatically created from the outliers.Finally,an empty cluster flag is checked to determine if some clusters need to be removed to ensure the favorable efficiency of subsequent clustering.The experimental results show that the proposed algorithm has good clustering accuracy and efficiency for both equal-length and unequal-length time series.关键词：time series;fuzzy clustering;adaptive incremental learning;outliers252|608|0更新时间：2025-07-16
- 摘要：Aiming at the problem that whale optimization algorithm is easy to fall into local extreme value and slow convergence speed,this paper proposes a whale optimization algorithm based on adaptive weight and simulated annealing.The improved convergence weight strategy is used to adjust the convergence speed of the algorithm,and the global optimization ability of the whale optimization algorithm is enhanced by simulated annealing.In the simulation experiment,18 test functions were calculated and the genetic algorithm,the particle swarm optimization algorithm and the standard whale algorithm were compared and statistically analyzed.At the same time,the influence of the adaptive weight and simulated annealing on the whale optimization is compared.The results show that the improved algorithm has a significant improvement in the calculation of the extremum of the test function,and the effectiveness of the improved algorithm is verified.关键词：intelligent optimization algorithm;whale optimization algorithm(WOA);adaptive weight;simulated annealing(SA)153|507|38更新时间：2025-07-16
- 摘要：To avoid the clustering results with the local scaling parameter σi of self-tuning may be influenced by outliers,and the unstable clustering results from K-means in self-tuning,two true self-adaptive spectral clustering algorithms were proposed.The two spectral clustering algorithms are respectively named as SC_SD(Spectral Clustering based on Standard Deviation) and SC_MD(Spectral Clustering based on Mean Distance).They respectively define the standard deviation of point i,and the mean distance from point i to others,as its radius of neighborhood,then count the number of points in the neighborhood,and use the standard deviation of point i in the neighborhood as its local scaling parameter,so as to avoid the influence from outliers to the local scaling parameter σi of point i,and the distortion in clustering results of self-tuning.SD_K-medoids are adopted to instead of K-means in self-tuning to avoid the unstable clustering results of K-means,so as to get the true clustering of a dataset.The experimental results on UCI datasets and on synthetic datasets demonstrate that SC_SD and SC_MD can obtain better clustering results than that of traditional spectral clustering algorithm NJW and spectral clustering algorithm self-tuning,and are robust to noises,and has got good scalability.The proposed SC_SD and SC_MD can detect the clustering of a dataset without any given information,and the SC_MD can be used to detect the clustering of a comparable big data.关键词：spectral clustering;neighborhood;standard deviation;mean distance;self-adaption335|414|14更新时间：2025-07-16
- 摘要：Sensor registration is the key precondition of the performance advantages of the multisensor data fusion system.In the presence of random errors,sensor biases,false alarms and missed detections,sensor registration usually works in a nonideal association envrionment.Traditional registration approches relying on the ideal association condition degrade seriously.On the other hand,traditional registration methods are sensitive to the target distribution.When targets are densely distributed,the registration problem is ill-conditioned and the estimation encounters the numerical instability phenomena.Focusing on sensor registration in the context of nonideal association and ill-condition,this paper presents the robust registration approach based on ridge least trimmed squares (RLTS).The proposed approach combines the advantages of the ridge regression (RR) and the least trimmed squares (LTS) estimation.The RLTS can deal with nonideal association and ill-condition simultaneously.Simulation results verify the robust performance of the RLTS method.关键词：sensor registration;system bias estimation;nonideal association;ill-condition;ridge least trimmed squares119|264|0更新时间：2025-07-16
- 摘要：Ladle is a container which used to connect the steel-making and pouring sections.The tapping temperature of the steel-making furnace is directly determined by the temperature of the inner wall of the ladle,which is closely related to the energy saving and consumption Reduction of the enterprise.Before the ladle is filled with molten steel,the ladle is placed in the hot repair station by driving.The absolute position of the ladle after placement is difficult to fix,resulting in a temperature field distribution that cannot be determined.So,this article presents a method for measuring the temperature field of ladle by laser positioning and scanning.Firstly,the positioning laser is used to determine the spatial coordinates of the ladle.Then,the actuator of the main measurement platform drives the scanning laser and the infrared temperature measurement sensor to scan the inner wall surface of the ladle to obtain the distance measurement and temperature of each measured point on the inner wall of the ladle.According to the spatial coordinate relationship,the distance measurement values of the measured are mapped to the ladle coordinate values,therefore we obtained the coordinates and temperature values of the inner wall of the ladle.So the temperature field distribution is obtained.The field experiment shows that the method can obtain the temperature field of the ladle,the coordinate positioning uncertainty is no more than 3.0mm;the maximum error of temperature measurement is 4.7℃,the minimum error is 0.5℃,and the average error is less than 3.3℃.The method achieves to high positioning and measurement accuracy,it will provide important information for steel-making control based on temperature information.关键词：infrared temperature measurement;ladle;inner wall temperature;laser positioning;scanning76|294|2更新时间：2025-07-16
- 摘要：In order to meet the requirements of grouping product families for advanced manufacturing modes such as mass customization,the features in BOM (Bill of Materials) are comprehensively analyzed,and a concept of BOM structure-based similarity metric model,a content-based similarity metric model,and an ensemble model combined with both are proposed.In the structure-based model,BOMs are represented by adjacent matrixes,including the relationships between materials and the quantity of materials,and the Orthogonal Procrustes Analysis is implemented to measure the similarity among BOMs.While in content-based model,effective text features are extracted from BOMs,being transformed to vectors by TFIDF(Term Frequency-Inverse Document Frequency),and finally being inputted into cosine approximation formula for similarity value.To obtain more accuracy and performance,a weight distribution method based on the Gini coefficient is proposed for the ensemble model.Finally,a test framework is provided and all models are in evaluated experimentally in accuracy and performance.关键词：similarity metric;BOM(Bill of Materials);product family;ensemble model73|306|0更新时间：2025-07-16
- 摘要：Currently,the DCA (dendritic cell algorithm) relies heavily on artificial experience to define the input signals in fault detection of different types of equipment,which is lack of adaptability and completeness.To address this problem,we propose a dendritic cell fault detection model based on numerical differentiation——NDDC-FD.In first place,according to change is the symptom and outward expression of system which is in danger,an adaptive signal extraction method based on danger perception of system status change is proposed,which uses numerical differentiation to calculate the change to extract the input signals.Next,the anomaly antigen evaluation method of original DC model can effectively detect abrupt fault,but it can't detect incipient fault in time.Therefore,the fault evaluation indicator based on concentration of T cells is proposed.Finally,our method is tested on DAMADICS and TE benchmark,and compared with DCA and PCA (principal component analysis).The results show that NDDC-FD method not only improves the adaptability of DCA,but also has higher detection rate than DCA and PCA,and has lower detection delay time in incipient fault detection.Overall,our method is generality and has well performance in the fault detection of industrial equipment.关键词：artificial immune systems;dendritic cells;numerical differentiation;fault detection81|298|2更新时间：2025-07-16
- 摘要：As Cloud Computing becomes an important information infrastructure,more and more applications are being migrated to the cloud.Therefore,the reliability of cloud services becomes increasingly important.In particular,the introduction of new edge computing mode puts forward higher requirements on the reliability of cloud services.How to guarantee the reliability of services through resource scheduling has become a hot topic of current research.In Cloud-Edge collaborative application scenarios,we research on a service reliability oriented cloud resource scheduling method to support cloud service reliability.And the cloud resource scheduling algorithm based on markov prediction model is put forward to solve the problem of task scheduling and load balancing in cloud service node failure situation,including the judgment of node load degree,the selection of migrated task and nodes,and the decision of migration routing.The goal is to achieve rapid cloud service recovery and to improve the reliability of cloud services.The experimental results show that the proposed method can effectively guarantee the service reliability.关键词：cloud service;reliability;resource scheduling;Markov process107|334|3更新时间：2025-07-16
- 摘要：Group delay is a very important indicator when measuring the impact of a transmission network on signal transmission time delay and signal distortion.The calculation of group delay based on the difference method is a commonly used in current measuring instruments,and there is a contradiction between frequency resolution and precision.It will inevitably lead to the decrease of measurement accuracy while improving the frequency resolution.Based on the analysis of the error source of the difference method,this paper presents a novel method of group delay measurement based on the Tikhonov regularization.The comparative analysis shows that the method can obtain the group delay value more precisely with higher frequency resolution when the measurement error is included.In the actual data calculation and verification,the validity of the method verified by comparison to the group delay data obtained by the vector network analyzer.关键词：group delay;difference method;Tikhonov regularization133|280|2更新时间：2025-07-16
- 摘要：Based on the Babin model and the sensitivity analysis of Evaporation duct on meteorological input parameters,a new diagnosis method——ensemble diagnosis (Babin_Ens) method——is proposed in this paper benefitting from the idea of weather ensemble prediction.By using this method,a comparison and analysis of the diagnosed result from the measurements on four island stations in the Chinese marginal seas is done.It is found that both root-mean-square error and bias of duct height and intensity for the Babin_Ens method are obviously improved compared to the Babin model,and the averaged bias improvements of duct height and intensity are 23.49% and 19.29% respectively.Numerical forecasts of evaporation duct are conducted by the Weather Research and Forecasting (WRF) model with the Babin_Ens method.The averaged bias improvement rates of duct height and intensity get improved respectively by 14.01% and 16.92%.It shows that the ensemble diagnosis method can greatly improve the diagnosis and prediction of evaporation duct.关键词：evaporation duct;Babin model;ensemble diagnosis;WRF model;numerical prediction119|429|2更新时间：2025-07-16
- 摘要：Aiming at that RGB image is rich in color details of scene and infrared image is sensitive to outline、size and boundary of target,a novel semantic segmentation model APFCN (Asymmetric Parallelism Fully Convolutional Networks) is proposed.In the upper part of APFCN,a five layer dilation convolution network,where the five kernel sizes are not uniform,is designed used to extract the high-level targets contour features of infrared image.In the lower part of APFCN,a classical CNN network is used to extract three scale features of RGB images.At the back of APFCN,the high level features of the infrared image are fused with the three scale features of the RGB image,and the fused features after 4 times upper sampling is used as the semantic segmentation output of APFCN.The results show that APFCN is better than FCN (input RGB image or infrared image) in PA (Pixel Accuracy) and MIoU (Mean Intersection over Union).APFCN is suitable for the semantic segmentation task of ground targets with consistent background colors.关键词：semantic segmentation;fully convolution neural network;asymmetric parallelism fully convolutional networks;dilation convolution;dilation rate195|497|8更新时间：2025-07-16
- 摘要：The impact of body effect on the total dose effect of ultra deep sub micron SOI devices is studied.130nm SOI NMOSFET devices were irradiated in TG bias state,the electrical parameters of devices before and after irradiation were monitored under different body bias.The short channel PD SOI NMOS devices are more sensitive to total dose radiation,and the larger channel width to length ratio of devices will cause more damage.After a certain dose of radiation,the partially depleted devices will turn into fully depleted devices,and the radiation-induced coupling effect can be observed.Significant threshold voltage shifts and large leakage currents can be observed after irradiation of 10 μ m/0.35 μ m devices.The transfer characteristic curve at negative body bias is positively shifted compared with it at zero body bias before the device is irradiated.When the body voltage Vb=-1.1V,then partially depleted devices become full-depletion devices.And continued increase in|Vb|will not result in continued increase in the width of the depletion area,which shows that the negative body bias can no longer adjust the width of the depletion region,so the transfer characteristic curve of the device does not show a positive shift similar to that before irradiation.关键词：total-ionizing-dose;silicon-on-insulator;body effect;shallow trench isolation81|303|2更新时间：2025-07-16
- 摘要：To solve the problem of the low detection accuracy of new intrusions with long detection time due to the complex and changeable nature of network intrusions,this paper proposes a network intrusion detection method based on the selective learning of Kernel Extreme Learning Machines (KELMs).First,based on the high efficiency learning characteristics of the single KELM learner,multiple KELMs are trained independently by the Bagging strategy.Then,based on the margin distance minimization (MDM) guidelines,KELM learners are integrated by selecting a part of them with high gains based on the MDM-based gain measures.Extensive validation and comparative experiments on the the KDD99 data set and on a hybrid network simulation platform mixed with wireless networks and Ethernet networks demonstrate that the proposed method achieves better recognition accuracies with faster recognition speed than the network intrusion detection methods based on the single learner and the traditional ensemble learning,which can effectively detect the known and unknown network intrusion connection in real time.关键词：network intrusion detection;extreme learning machine (ELM);anomaly detection;selective ensemble learning;margin distance minimization198|350|10更新时间：2025-07-16
- 摘要：We investigate in this paper how to effectively reduce the risk of privacy leakage caused by refactoring attacks when the sensitive attributes and some quasi-identifier attributes are correlated.Firstly,the mutual information theory is used to find the quasi-identifier attributes which have strong dependence on the sensitive attributes in the original dataset,which provides a theoretical basis for accurately perturbing the data attributes.Secondly,for the correlated attributes and the non-correlated attributes,the invariant random response method is applied to perturb a certain data attribute or a combination of data attributes to satisfy the local ε-differential privacy requirement.Theoretical analysis of the impact of data perturbations on privacy leakage probability and data utility is also conducted.Finally,the experiment verifies the effectiveness of the proposed algorithm and its ability to process incremental data.The experimental results demonstrate that the algorithm can achieve a better balance between data utility and privacy protection.关键词：local differential privacy;invariant post-random response;data reconstruction;data perturbation;privacy protection162|364|5更新时间：2025-07-16
- 摘要：Clustering a large number of short texts in social media has great value in applications.However,short texts often have these characteristics:lots of noises,growing rapidly and massive data.Most existing short text clustering algorithms are not effectively enough to process such short texts.Aiming at this problem,we propose an algorithm of short text online clustering based on incremental robust nonnegative matrix factorization (STOCIRNMF).This algorithm uses NMF to build the short text clustering model and applies l2,1 norm to devise its objective function for improving its robustness.Meanwhile,STOCIRNMF can cluster short texts incrementally by using incremental iterative update rules.We conduct extensive experiments on real Sohu news titles and Weibo datasets.The results show that STOCIRNMF not only has better performance of short text clustering than some representative algorithms,but also is very effective to detect micro blog's topics online.关键词：short text clustering;robust nonnegative matrix factorization;online clustering;l2,1 norm;incremental iterative update rules74|281|2更新时间：2025-07-16
- 摘要：In the past decades,the amount of information data is growing in an unexpected speed and how to store and protect these large amounts of data becomes a problem.Cloud storage and data deduplication technology are the principal technology to solve the above problem.Deduplication technology is widely used in cloud storage systems.However,there are some shortcomings such as the expansion of the index information and the uncertainty of the data block in the current mainstream cloud storage technology,which lead to the waste of space and a big difference in the length of the data block.To overcome these shortcomings,through the study of cloud storage and data deduplication,this paper presents an optimal hierarchical deduplication strategy based on content defined chunking and proposes the corresponding algorithm,and achieves the purpose of saving memory space and obtaining better compression performance.Finally,this paper selects the content of the CNN news as a test set.By comparing the compression ratio and the compression time,the optimal hierarchical deduplication strategy has increased compression ratio by 3% and reduced compression time by 2% compared with the current mainstream deduplication strategy.关键词：cloud storage;data deduplication;data block partition;hierarchical;compression ratio88|323|1更新时间：2025-07-16
- 摘要：Model-based diagnosis is an important branch of research in the field of artificial intelligence.The efficiency for generating all minimal hitting sets,i.e.,candidate diagnoses,considerably affects the final diagnostic process.This paper focuses on the current major algorithms for computing minimal hitting sets.First,the basic ideas of algorithms were briefly introduced.Then,the similarities and differences,and complexity of them were compared by simple algorithm description and examples.An integrated experimental platform was implemented for testing and comparing their time efficiency,which provides an important reference for the actual selection of an appropriate algorithm in practice.关键词：model-based diagnosis;hitting set;performance167|307|4更新时间：2025-07-16
- 摘要：Attribute reduction is one of important applications in rough set theory.There are more than one attribute reduct in a data set,and heuristic algorithms are always used to find one of them,which is verified with experiments.For many attribute reducts,it is hard for people to distinguish them,and lacks of valid methods of selecting the best one or a better one.Indexes of concept drift and information loss are employed to compare the same type of Pawlak attribute reducts in a knowledge system.The focus of attribute reducts is presented,and its properties are investigated in this paper.Experimental results show that the closest attribute reduct to the focus of attribute reducts is better than other attribute reducts in classification accuracy.Indexes of concept drift detection and information loss can distinguish different attribute reducts,and the focus of attribute reducts can be employed to select the best attribute reduct or a better one.关键词：rough sets;attribute reduction;concept drift;focus of attribute reducts82|259|0更新时间：2025-07-16
- 摘要：Keyphrases are used to summarize the document and high-quality keyphrases have great importance in text summarizing,reading and indexing.However,most studies of keyphrase extraction have strict limitation in the form of patterns,and are unable to achieve the semantic relation between words and phrases.The results are failure to autonomously extract keyphrases.Keyphrase extraction using sequential patterns mining with one-off and general gaps condition algorithm (KEING) is proposed in this paper.Taking into account one off condition and general gaps,SPING(Sequential Patterns mIning with oNe-off and General gaps condition)can catch semantic relations between words and phrases more effectively.Therefore,KEING will get effective candidate keyphrases and count their features.Then a supervised machine learning method is used to train features and construct a classification model,we can extract keyphrase with this model.Experimental results demonstrate KEING can effectively extract high quality keyphrases.关键词：general gap;sequential patterns mining;keyphrase extraction;machine learning170|353|1更新时间：2025-07-16
- 摘要：To resolve the crosstalk problem of the transmission lines in high-speed interconnection system,the properties of the transmission matrix are studied.Both the model of the transmission matrix and the matrix decomposition form are established in the following two cases,one is the crosstalk between the disturbed line and its one adjacent microstrip line on both sides,and the other is the crosstalk between the disturbed line and the two adjacent lines on both sides.The crosstalk cancellation scheme for the above two cases is given,and it is verified with the advanced design system (ADS).The results show that both signal jitter and distortion are greatly improved,the crosstalk is obviously cancelled,and it is illustrated that the crosstalk cancellation effect of the scheme considering two adjacent microstrip lines is better than that of one adjacent microstrip line.It is practical to maintain the integrity of the high-speed signal.关键词：crosstalk;coupled transmission lines-channel transmission matrix(CTL-CTM);microstrip lines75|310|3更新时间：2025-07-16
- 摘要：Bandpass filters (BPFs) with generalized Chebyshev response using dual-and single-mode substrate integrated waveguide (SIW) resonators are proposed according to the miniaturization and high performances.The box-like topology with an additional cross-coupled path realized by the fundamental mode (TE101) of the dual-mode SIW cavity is proposed,and an additional transmission zero (TZ) is achieved.The traditional negative coupling structure is not required in the proposed structure,and two TZs can be realized;moreover,the structure has flexible responses that two TZs can be located on the lower and/or upper of passband.To improve the selectivity,a complementary split-ring resonator (CSRR) etched into the fourth-order SIW filter to design a filter with an extended box-like topology is studied,and the fifth-order filtering function with three finite TZs can be realized.To verify the feasibility,a symmetric and an asymmetric response fourth-order BPFs with center frequency of 10GHz and a fifth-order BPF with center frequency of 5.8GHz are designed.The corresponding coupling matrixes of box-like topologies with nonresonating node (NRN) are also given to verify the proposed structure.Finally,the designed SIW BPFs are processed and measured.The responses of coupling matrixes,simulated and measured results are in good agreement,which demonstrates the feasibility of designed filters with high performances using the proposed structures.关键词：bandpass filter (BPF);nonresonating node (NRN);box-like topology;substrate integrated waveguide (SIW);generalized Chebyshev;asymmetric response;complementary split-ring resonator (CSRR)151|376|3更新时间：2025-07-16
- 摘要：Software-Defined Networking (SDN) separates data plane from control plane,which makes it more flexible,opening and programmable,compared with traditional IP networks.However,the separation incurs many security problems.In this paper,we find that we can construct controller-to-switch loop (CSL) attacks by leveraging dedicated rules and well constructed packets.The attacks can effectively exhaust controller resource,which leads to denial of service (DoS).The existing OpenFlow policy verification schemes only focus on detecting data plane loop,and cannot detect such controller-to-switch loops.In order to detect CSL attacks,we proposed a novel policy verification scheme.The scheme constructs a packet forwarding graph by analyzing network update events and packet-out messages,and efficiently identifies the forwarding loops by traversing the graph.In order to evaluate our defense,we implement it in the Floodlight controller,and perform experiments with Mininet.The experimental results show that our defense can precisely detect the loop attacks and effectively throttle them.关键词：software-defined networking;control plane;data plane;controller-to-switch loop detection;policy consistency check96|292|0更新时间：2025-07-16
- 摘要：In recent years,deep learning based scene text detection have achieved significant progress.The paper reviews state-of-the-art methods in the field from 2014-2018.We categorize existing methods into traditional Region Proposal based method,Text Proposal Network method,segmentation based method and hybrid method based on Text Proposal Network and segmentation with detailed analysis of pros and cons for the four methods.Finally,we point out research trends and focuses in this field.关键词：deep learning;scene text;text detection257|572|12更新时间：2025-07-16
- 摘要：Human action recognition in videos is a challenging topic in the field of computer vision.It is widely not only used in video information retrieval,daily life security,public video surveillance,but also human-computer interaction,scientific cognition and other fields.First,the research background,research significance and difficulties of action recognition are briefly introduced,and then the deep learning model based action recognition methods are comprehensively reviewed from three different aspects:the types and numbers of input signals,the combination with traditional feature extraction methods,and the pre-trained datasets.Furthermore,the performances of some typical methods on UCF101 and HMDB51 datasets are overviewed and analyzed.Last the possible future research directions are discussed from three perspectives:the video data preprocessing,the video human motion feature representation,and the model training.关键词：action recognition;review;convolutional neural network;deep learning463|1238|28更新时间：2025-07-16
- 摘要：In this paper,based on lattice evaluation theory and by defining probality measure in pre-rough algebra evaluation lattice and set of all formulae respectively,the new type of rough probabilistic truth degree of formulae in rough logic is introduced by the integral method.The MP rule,HS rule and meet inference of rough probabilistic truth degree are proved,the concept of accuracy degree and roughness degree of fomulae are introduced also.At the meantime,the concept of rough similarity degree and pseudo-distances between formulae are introduced and three different kinds of approximate reasoning models are estabished.The theory of quantitative logic is expanded to rough logic,which makes it possible in graded reasoning in rough logic.关键词：rough logic;pre-rough algebra;rough probabilistic truth degree;rough similarity degree;approximate reasoning88|240|0更新时间：2025-07-16
- 摘要：With the advent of the era of big data,there is an increasingly urgent need for ultra-high precision scientific calculation.One of the difficulties is the large number factorial problem.Stirling formula as the traditional approximate method of calculating factorial of large numbers cannot meet the accuracy requirements.Other factorial algorithm can achieve high accuracy,but at the cost of a large amount of storage space.This paper puts forward a factorial algorithm of large numbers with zero error that can optimize storage space according to the problem size.Besides,the computing speed can be improved by making full use of parallel computing and the advantages of field programmable gate array (FPGA).The test results show that the proposed algorithm has better efficiency of time and space,which can be used in many fields such as large factorial calculator.关键词：factorial algorithm;storage space;parallel computation;FPGA (field programmable gate array)225|447|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