Author Login Editor-in-Chief Peer Review Editor Work Office Work

15 June 2017, Volume 43 Issue 6
    

  • Select all
    |
  • LU Siyu,WANG Hongwei,ZHANG Youhui,YANG Guangwen,ZHENG Weimin
    Computer Engineering. 2017, 43(6): 1-10,18. https://doi.org/10.3969/j.issn.1000-3428.2017.06.001
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    According to the principle of computations approaching data,this paper proposes a design method of Network-on-Chip(NoC) oriented to MPI collective operations,which focuses on the hardware enhancement of common NoC routers to speed up MPI collective operations on the network layer.It designs MPI reduction,extends it to support more operations and combines it with an adaptive method for the deterministic routing algorithm,which can learn transmission paths of messages dynamically.Thus,enhanced routers can complete message processing in place,which not only speed up the processing procedure but also coalesce messages.The design method for detailed micro-architecture of NoC is presented.Different layout strategies of enhanced routers are compared and the corresponding performance,power consumption and extra chip-area are evaluated.Testing results show that,compared with ideal software-based implementation,the proposed method can improve the reduction performance by 6.4~41.7 times,broadcast by 15.3~31.2,global reduction by 5.4~9.7 times,and gather by 1.3~1.8 times,while the consumption of power and chip-area is limited.
  • LUO Hui,YING Shi,LI Lin,DONG Bo
    Computer Engineering. 2017, 43(6): 11-18. https://doi.org/10.3969/j.issn.1000-3428.2017.06.002
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The existing Software as a Service(SaaS) dynamic deployment models and languages mainly focus on deployment configuration information and performance or resource availability requirement when describing software deployment scheme,which cannot support performance optimization of SaaS software deployment scheme.Based on the above problems,this paper puts forward a SaaS Software Deployment Language(SSDL) according to SaaS software deployment’s characteristics and information SaaS software performance optimization requires.This paper constructs the conceptual model of SSDL describing SaaS software deployment scheme from four aspects which are deployment environment,deployment object,deployment plan and deployment constraint,and then presents the performance optimization method based on SSDL.Finally,a deployment scheme performance optimization experiment of an emergency task operation management system verifies the support function of the proposed SSDL for software performance optimization.
  • REN Shengbing,ZHANG Jianwei,WU Bin,WANG Zhijian
    Computer Engineering. 2017, 43(6): 19-23,29. https://doi.org/10.3969/j.issn.1000-3428.2017.06.003
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The credibility of program verification results and the verification efficiency in Satisfiablity Modulo Theories(SMT)-based bounded model checking are influenced greatly by bounds.However,the traditional serial checking method cannot validate under the conditions of too large bounds because of the limitation of handling performance and memory in a single machine.In order to solve this problem,this paper proposes a SMT-based distributed BMC method in Spark.First of all,the LLVM Intermediate Representation(LLVM-IR) translated from the source program is converted into Spark built-in data structure Pair Resilient Distributed Dataset(RDD).Afterwards,the Pair RDD is converted into Verification Conditions RDD(VCs RDD) which is then converted into SMT-LIB with the proposed MapReduce algorithm.In the end,the proposed method utilizes SMT solver for verification with the SMT-LIB.Experimental results indicate that,compared with the traditional serial checking method,the proposed method improves not only the bounds of the validation process and the correctness of the verification results,but also the speed of verification for the program with higher complexity under the same bound.
  • LI Shuangquan,DU Yajuan,LIU Lingfeng
    Computer Engineering. 2017, 43(6): 24-29. https://doi.org/10.3969/j.issn.1000-3428.2017.06.004
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to implement the rendezvous and docking of two space aircrafts,the simulation debugging of the rendezvous and docking process must be done.So this paper builds a space docking hardware-in-the-loop simulation test-bed which overcomes the limit of no physical simulation environment.It can reproduce the rendezvous and docking process of two space aircrafts.The design scheme of measurement and control system for docking mechanism is put forward.Its software structure and control flow process are described,and the measurement and control model of docking mechanism is built.The model is downloaded to run through the RT-LAB real time simulation platform,and implements sensor signal sampling and motion control of docking mechanism.Experimental results show that the system model can simulate rendezvous and docking process of aircrafts.It has good real-time performance,and the system runs stably and reliably.
  • ZHANG Jing,WANG Liang,FAN Hongbo,XIAO Zhibin
    Computer Engineering. 2017, 43(6): 30-34,39. https://doi.org/10.3969/j.issn.1000-3428.2017.06.005
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The existing software energy consumption analysis methods mainly research the energy consumption on the algorithm level and the instruction level,rarely on the software architecture level.From a software architecture for embedded systems,value and time tags are added on the basis of traditional Petri nets,so as to establish a Value Time Petri Net(VTPN).The software component and linker energy consumption models are established through VTPN to analyze the system energy consumption,and the minimum energy consumption path algorithm is proposed to reduce the system energy consumption.Experimental results show that the actual energy consumption results obatined by using the proposed models are close to the expected results.The model can reduce system overall energy consumption and improve resource utilization.
  • WANG Qiangjun,LIU Guofan
    Computer Engineering. 2017, 43(6): 35-39. https://doi.org/10.3969/j.issn.1000-3428.2017.06.006
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Aiming at the problem of unequal energy consumption of nodes and short period of network stability in the operation process of Wireless Sensor Network(WSN),an improved Stable Election Protocol(SEP) is proposed.The cluster head election mechanism which comprehensively considers node residual energy and Energy Consumption Rate(ECR) improves the node threshold calculation,making the nodes with more residual energy and less energy consumption rate more likely to become cluster head nodes and thus avoiding low-energy nodes assuming more network load and balancing energy consumption of nodes.Simulation results illustrate that the presented protocol can effectively reduce the network energy consumption and prolong the network life.
  • YANG Mi,CHEN Jianzhong,NIU Yingtao
    Computer Engineering. 2017, 43(6): 40-45,52. https://doi.org/10.3969/j.issn.1000-3428.2017.06.007
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    To deal with the uncertainty description of index in the evaluation of communication electronic defense effectiveness,an information fusion evaluation method based on cloud model and evidence theory is proposed.A cloud generator of comment levels is designed,and the Basic Probability Assignment(BPA) function is constructed by cloud membership grade.The BPA information of different indexes is synthesized based on the evidence fusion rules,and the final evaluation result is obtained through linearly weighting the synthetic BPA and quantization values of the comment levels.Taking the frequency hopping communication equipment as an example to evaluate communication electronic defense effectiveness,the results show that the method can effectively reduce the subjectivity caused by expert evaluation and make the evaluation result more accurate and objective.
  • GUO Xiaobo,LI Songtao,ZHANG Dexian
    Computer Engineering. 2017, 43(6): 46-52. https://doi.org/10.3969/j.issn.1000-3428.2017.06.008
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to overcome the problem of serious signal interference induced by Internet of Things(IoT)node data confliction in current IoT node anti-collision algorithm,a new IoT node information anti-collision algorithm is proposed.The discrete orthogonal modulation and demodulation mechanisms of the radio frequency identification signal are constructed to reduce the channel collision caused by the mutual interference among the nodes.Combining backoff power perception mechanism is built and node data collision avoidance rules are defined to improve the fluctuation range of the transmitted power.Experimental results show that this algorithm has a longer signal recognition distance and lower channel collision probability compared with ColorWava and Decision Information Controlling Algorithm(DICA).
  • ZHANG Huan,LIU Yan,TONG Baotong,ZHAO Lingxiao,YANG Yingxue,WANG Yuping,DAI Yakang
    Computer Engineering. 2017, 43(6): 53-58. https://doi.org/10.3969/j.issn.1000-3428.2017.06.009
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Aiming at the problem of modal selection bias in Complete Ensemble Empirical Mode Decomposition(CEEMD),this paper proposes a new Electroencephalogram(EEG) signal denoising method by combining improved CEEMD(ICEEMD).First,the EEG signal is decomposed to several Intrinsic Mode Functions(IMF) by ICEEMD.Then,the approximate entropy of each IMF is calculated respectively.Finally,the IMF with the maximum approximate entropy is chosen as the denoised result.The experiments result based on analog signals and real EEG signals shows that,compared with Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(ECCMDAN),the new method can give more clear and stable denoising results,and it also solves the problems such as inaccurate denoising and false mode caused by the blind selection of IMF.
  • (西安财经学院 实验实训教学管理中心,西安 710061)
    Computer Engineering. 2017, 43(6): 59-65. https://doi.org/10.3969/j.issn.1000-3428.2017.06.010
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to solve the problem of large amount of computation and difficulty in the optimization of the existing Reader Network Planning(RNP),this paper proposes a new constrained method of Radio Frequency Identification(RFID) network planning based on micro Genetic Algorithm(GA),so as to optimize the RFID reader.The method of spatial intersection is used to solve the problem of searching ability degradation caused by a small number of population.In order to reduce the iteration cost,the correction scheme includes the elimination of redundant reader to make the feasibility of chromosome to be low-cost chromosome after correction.Experimental results show that compared with Canonical Genetic Algorithm(CGA) and Particle Swarm Optimization(PSO)methods,the proposed method has high computational efficiency,and has a great advantage in both computation time and label coverage.
  • SONG Gaojun,HU Cheng,ZHOU Fang
    Computer Engineering. 2017, 43(6): 66-72. https://doi.org/10.3969/j.issn.1000-3428.2017.06.011
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Considering that the traditional MPLS-VPN network architecture has problems like inadequate scalability,low security and poor transmission quality,a method using hierarchical Provider Edge(PE) technology to optimize network architecture is proposed.The method uses the hierarchical way to optimize the network topology and achieves the maximum separation of the routing control and data forwarding.The weight of the link utilization is introduced to optimize Equal-Cost Multipath(ECMP) selection algorithm and various performance indicators of the network are compared before and after the optimization.Experimental results show that this method can make the optimized traffic load more balanced and significantly reduce packet delay,jitter and packet loss rate,and it obviously improves network scalability,security and transmission quality.
  • PENG Yanbin,TIAN Ye,PENG Xinguang
    Computer Engineering. 2017, 43(6): 73-77. https://doi.org/10.3969/j.issn.1000-3428.2017.06.012
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    For the information transmission in medical Wireless Body Area Network(WBAN),authentication mechanisms based on biological features and public key infrastructure with trusted third party have drawbacks of identity false negative,false positive and high deployment cost.On the basis of analyzing security and performance requirement of authentication mechanism for medical WBAN,Ecliptic Curve Cryptosystem(ECC) encryption and signature algorithm is introduced into existing lightweight authentication protocols based on challenge and response.Based on the best trade-off perspective among experience,performance,security and cost,an end-to-end lightweight authentication protocol is proposed.Theory proof and experiment results indicate that the protocol is provided with superiority of end-to-end,light weight,mutuality,unlinkability,low cost and so on.It is an excellent candidate to medical WBAN.
  • LU Yu,WANG Xueming
    Computer Engineering. 2017, 43(6): 78-83,91. https://doi.org/10.3969/j.issn.1000-3428.2017.06.013
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to improve the computing efficiency of scalar multiplication on hyperelliptic curves,the skew-Frobenius mapping on the elliptic curve is extended to hyperelliptic curves,and a skew-Frobenius mapping is constructed on the hyperelliptic curve with a genus of 4,through the skew-Frobenius mapping on the hyperelliptic curve which is defined on genus 2,3 and 4.The general form of skew-Frobenius mapping on the hyperelliptic curve is proposed.A new scalar multiplication algorithm is constructed based on the genetal form of skew-Frobenius mapping on hyperelliptic curves,which improve the efficiency of scalar multiplication on hyperelliptic curves.Experimental results show the scalar multiplication of the skew-Frobenius mapping on the hyperelliptic curve is about 39% faster than that of the binary scalar multiplication algorithm.
  • DUAN Xiaoyi,SHE Gaojian,GAO Xianwei,HE Siman,CUI Qi,WANG Sixiang,MAO Zeyu
    Computer Engineering. 2017, 43(6): 84-91. https://doi.org/10.3969/j.issn.1000-3428.2017.06.014
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In Correlation Power Analysis(CPA) attack,the existence of power curves noise affects the success rate of attack results.For this reason,this paper proposes a new method of power curve preprocessing using wavelet packet threshold.The wavelet packet threshold method is first used to denoise the power curves and then perform the correlation power analysis.Experimental results show that wavelet packet threshold denoising can remove irrelevant noise in the power curves,improve the Signal to Noise Ratio(SNR) of the useful signal,reduce the data needed for the key analysis and improve the efficiency of the CPA.
  • WANG Jianxin,WANG Boren,QU Ming,ZHANG Lei
    Computer Engineering. 2017, 43(6): 92-96. https://doi.org/10.3969/j.issn.1000-3428.2017.06.015
    Abstract ( ) Download PDF ( )   Knowledge map   Save

    The accuracy of the hardware Trojan detection method based on the traditional Euclidean distance discrimination is too low to meet the demand.For this reason,the side-channel power consumption information is collected and analized in this paper,and an exponential transformation optimization scheme is given based on the extra power consumption caused by the hardware Trojan module.This adjustable parameter is able to increase the recognition of the hardware Trojan.Experimental results show that compared with the traditional Euclidean distance discrimination method,the modified Euclidean distance scheme with adjustable parameters improves the discrimination performance by 29% and the Trojan detection accuracy reaches 98%.

  • LI Zichen,LIU Boya,WANG Peidong,YANG Yatao
    Computer Engineering. 2017, 43(6): 97-100,104. https://doi.org/10.3969/j.issn.1000-3428.2017.06.016
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    To ensure the wireless communication security of the reader and tag in the Radio Frequency Identification(RFID) system,based on domestic public key cryptography algorithm SM2 and introducing zero knowledge proof idea,this paper proposes a two-way authentication protocol.Then security and efficiency analysis are given.The formal analysis is made by using BAN logic.Results show that the two-way authentication of the proposed protocol is completed in the case that reader and tag just interact twice,and has high safety and communication efficiency.
  • CHEN Shiwei
    Computer Engineering. 2017, 43(6): 101-104. https://doi.org/10.3969/j.issn.1000-3428.2017.06.017
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The cryptographic algorithm including modular addition operation(Addition),rotation shift operation(Rotation) and XOR operation(XOR) is called ARX type algorithm.The mixed use of the 3 operations can achieve better diffusion and confusion effects.This paper defines the ARX function of two variables,researches on the simultaneous linearization condition of two iterations of the function.Using the method of statistic analysis,the computational formula of the element number for guaranteeing the linearization conditions is given.Analyzing the probability of the linearized condition establishment under the condition that the two iterations are independent,it is found that the method of statistic analysis more accurately describes the factors that affect the probability of the linearized condition.Moreover,the addition of a left and right block transform does not affect the linearization condition of the ARX function.
  • MAO Qianren,WANG Chaobin,BAO Xiao’an,LI Yanmei
    Computer Engineering. 2017, 43(6): 105-110,117. https://doi.org/10.3969/j.issn.1000-3428.2017.06.018
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to explore the evolution law and guidance effect of Internet public opinion,a public opinion guidance mechanism based on Cellular Automaton(CA) model is proposed using the quantitative analysis method.This paper defines some subject attributes which are the internal motivation coefficient,the stubborn strength coefficient,etc.and puts forward a guidance coefficient of public opinion.The subject attributes and the guidance strength coefficient are assigned to each unit cell in CA model.The effect of external guidance and the characteristics of netizen on the intensity and direction of the public opinion development are simulated and analysed.Experimental results show that,at the formation and expansion stages of network public opinion dissemination,with a correct guidance,the outbreak of bad public opinion can be avoided,and the evolution of public opinion generally presents a diffusion state.
  • WANG Yue,LIU Yaqiu,GUO Jifeng,JING Weipeng
    Computer Engineering. 2017, 43(6): 111-117. https://doi.org/10.3969/j.issn.1000-3428.2017.06.019
    Abstract ( ) Download PDF ( )   Knowledge map   Save

    In order to meet the varied Quality of Service(QoS) requirements of user tasks under cloud computing environment,taking into account of task deadline,scheduling budget and reliability,this paper proposes a task scheduling algorithm with multi-QoS constraints based on Discrete Particle Swarm Optimization(QoS-DPSO).Firstly,the QoS of the task is defined and its mathematical model is established.Then,the search space of DPSO is constrained according to the deadline and scheduling budget,and the fitness function of DPSO is redefined with reliability.Finally,according to the fitness value,the optimal task scheduling scheme is searched out.Experimental result shows that,compared with PSO,DPSO,DBC and EDF algorithm,QoS-DPSO has higher reliability in the case of meeting the deadline.Meanwhile,it has less influence on the performance of Makespan.

  • WANG Hui,YU Bo,HONG Yu,XIAO Yanghua
    Computer Engineering. 2017, 43(6): 118-124. https://doi.org/10.3969/j.issn.1000-3428.2017.06.020
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to effectively extract huge amounts of Web information in multiple fields,a Web information extraction system is designed based on Chinese knowledge graph,CN-DBpedia.Firstly,webpage data items with noise are automatically labeled based on knowledge graph.Then,correct wrappers are induced and learned from labeling sets with errors by a fault-tolerant wrapper induction framework.Experimental results demonstrate that,compared with traditional information extraction method by manual annotation,the proposed system has higher precision and recall rate.It can significantly reduce human participation during the extraction process and flexibly apply to large-scale webpage information extraction tasks in multiple fields.
  • TAO Zhao,YANG Jianbo,ZHANG Bo,ZHANG Liyun
    Computer Engineering. 2017, 43(6): 125-128,135. https://doi.org/10.3969/j.issn.1000-3428.2017.06.021
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In the pattern matching of bit stream,because the character sets of target string and pattern string are simple,the average jumping length of matching window is short during the matching process,making Quick Search(QS) matching not efficient.So this paper analyses the relationship between the matching efficiency of bad character heuristic rules and the character set size of QS algorithm.On the reference of coding QS algorithm,it proposes a grouping QS algorithm for group preprocessing of pattern strings and using the character set to calculate the jump set,and presents a method to calculate the bad character group heuristic rule and the optimal group length.Experimental results show that compared with the non-grouping algorithm,the proposed algorithm can improve the average hop length of the matching window during the matching of bit stream pattern strings,and the computational efficiency is better.
  • LIU Huicong,TIAN Bingjie,TIAN Xuedong
    Computer Engineering. 2017, 43(6): 129-135. https://doi.org/10.3969/j.issn.1000-3428.2017.06.022
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Aiming at the difficulties in achieving retrieval that result from the diversity of the mathematical expression structure,a method of mathematical expression indexing and retrieval is proposed.Through analysis and induction of LaTeX mathematical expression’s characteristics,a mathematical expression feature representation way is defined for the two-dimensional structure characteristic in the indexing stage.And the inter-relevant successive tree indexing pattern is applied to the construction of the mathematical expression indexing,so as to solve the problem of the hierarchical growth of the tree structure representation.In the matching stage,the matching algorithm of query pattern which includes exact matching,compatible matching,sub-expression matching and fuzzy matching is designed.In the browser/server mode,51 076 mathematical expressions are used in the experiment of indexing and matching.The results show the designed indexing and retrieval method accelerates the query speed and reduces the storage space,which can adapt the structure characteristics of the mathematical expression and achieve better retrieval effect.
  • LU Jing,SONG Bin,XIANG Wanhong,WU Shihong,SUN Xiaodong,TANG Jing
    Computer Engineering. 2017, 43(6): 136-144. https://doi.org/10.3969/j.issn.1000-3428.2017.06.023
    Abstract ( ) Download PDF ( )   Knowledge map   Save

    In order to realize the automatic electronization for source document,this paper proposes an automatic source document classification method based on image features.Based on the document layout analysis of rule hypothesis tree,an intelligent document-filling system with source document electronization is designed,which can be applied to mobile phones,pads and other mobile devices.During online reimbursement,document information can be automatically input to the financial information system merely by scanning the source document with the mobile device.It can not only eliminate the waste of time and resources caused by human input,but also ensure that the source document and each business account during online reimbursement are consistent.Experimental results show that the document classification accuracy of this method is 88.38%.The accuracy of intelligent document-filling is 87.22%,and it takes 5.042 s to deal with a document on average.The proposed method has high economic value and application value.

  • SONG Chao,XU Daoyun,QIN Yongbin
    Computer Engineering. 2017, 43(6): 145-149,157. https://doi.org/10.3969/j.issn.1000-3428.2017.06.024
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    As the Convolutional Neural Network(CNN) mainly uses the local features of the image,ignoring image channel features,this paper proposes a Detached Multiple Convolutional Neural Network(DMCNN).It extracts the channel features and convolution features,and fuses them in the whole connection layer so that image recognition and classification effects of the proposed network are improved.The experimental results on cifar10 and SVHN datasets show that the average recognition rate of the network is higher than that of other 8 CNNs like RexNet,Network in Network, Maxout.
  • ZHANG Zhiyuan,YANG Hongjing,ZHAO Yue
    Computer Engineering. 2017, 43(6): 150-157. https://doi.org/10.3969/j.issn.1000-3428.2017.06.025
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Mining the probability distribution of topic words in document collection can make a summary understanding of the document content.Further exploring the connection relationship between words in a given topic not only riches the meaning of topic words,but also reveals the hierarchy and aggregation of topics.For the labeled document collection,this paper proposes a method to compute the conditional probability of two words under a given topic based on Gibbs sampling outputs of labeled Latent Dirichlet Allocation(LDA),and a topical text network is also constructed.Compared with Pointwise Labeled-LDA(PL-LDA) model,this method does not extend the original document and needs less computation cost and shorter time.Experiments on the data set of aviation safety reports show that,for topics with many labeled words,this method can better display the distribution of subject words and the complex relationship between them.
  • WU Qiong,ZENG Qingpeng
    Computer Engineering. 2017, 43(6): 158-168. https://doi.org/10.3969/j.issn.1000-3428.2017.06.026
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    To obtain positive and negative quantitative association rules simultaneously and reduce the impact of human intervention as far as possible,a mining algorithm of positive and negative quantitative association rules based on multi-objective firework optimization algorithm is proposed.It introduces comprehensive search association rules and uses external library to store non-dominated solution.The diversity of association rules in the library is maintained by means of the redundancy elimination mechanism based on similarity.After several iterations,the association rule set is obtained.Results of experiments on real dataset show that there is no need for the proposed algorithm to specify the thresholds of support or confidence,and both positive and negative quantitative association rules can be mined in only one single step.In contrast with Apriori algorithm and single objective evolutionary algorithms,this algorithm gives stable results upon different real-world datasets,and the dataset can be fully covered.At the same time,a better trade-off between reliability,relevance and comprehensibility is achieved.
  • YANG Fengqin,SONG Meijia,SUN Tieli,SUN Hongguang
    Computer Engineering. 2017, 43(6): 169-176. https://doi.org/10.3969/j.issn.1000-3428.2017.06.027
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Traditional opinion target extraction methods ignore the opinion targets with low frequency,complicated structures and large amount of information.Aiming at this problem,an integrated opinion target extraction method for Chinese product reviews is proposed.A basic word unit set is obtained according to part-of-speech and syntax,and it is extended based on its integrity value and deficiency value.A concatenation stability rule is applied to decide whether the extension is reasonable.After above processing,a candidate opinion target set is obtained.The final opinion target set is the result after some unreasonable candidate opinion targets are filtered from the candidate set by filtering strategies.The experimental results on real data sets show that the proposed method works well for the opinion targets with low frequency and complicated structures.
  • LI Xiaohong,CAO Lin,SU Yun,MA Huifang
    Computer Engineering. 2017, 43(6): 177-181. https://doi.org/10.3969/j.issn.1000-3428.2017.06.028
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    By analyzing high dimension characteristic and sparsity of short text,this paper proposes a feature extension algorithm fusing statistical information feature words between concepts and semantic similarity for short text.Firstly,it selects reasonable feature set through the contribution degree of word and constructs initial feature extension set.Then it calculates statistical correlation between feature words and constructs a binary word correlation pair set.Finally,by using the semantic relations of external knowledge base,HowNet,it obtains synsets of relevant words,calculates the semantic similarity,extends the synsets which meet the conditions to the feature words of the short text and obtains the extend feature set.Experimental results show that,after using the proposed algorithm to extended features,the classification results of classifiers can be greatly improved.
  • XIAO Bao,LI Pu,HU Jiaojiao,JIANG Yuncheng
    Computer Engineering. 2017, 43(6): 182-188,194. https://doi.org/10.3969/j.issn.1000-3428.2017.06.029
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The characteristics of microblog such as short text,sparse feature and the semantic gap between users’ query may reduce semantic retrieval efficiency.Aiming at these problems,taking into account both text feature and semantic of knowledge base,a semantic retrieval model based on latent semantics and graph structure is proposed.Firstly,Tversky algorithm is employed to measure feature relatedness by taking Hashtag as feature;Secondly,a topic model is trained by Latent Dirichlet Allocation(LDA)for Wikipedia,and text topic relatedness mapped to this model is calculated by JSD;Finally,the connection graph of entity and its network relation are extracted in DBpedia.SimRank is employed to measure relatedness between two entities.The three types of relatednesses calculated in previous steps are used to compute a final relatedness.Twitter subsets for short and long queries are used in experiment.Experimental results show that,compared with the method based on linked open data and graph-based theory,the proposed model improves MAP,P@30,R-Prec by 2.98%,6.40%,5.16% respectively,which means that it has better retrieval perfermance.
  • DING Wei,ZHOU Feng,MIAO Jiepu,YU Zhengtao,ZHOU Lanjiang,YAN Xin
    Computer Engineering. 2017, 43(6): 189-194. https://doi.org/10.3969/j.issn.1000-3428.2017.06.030
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    For the temporal relation recognition task in Chinese corpus provided by TempEval-2010 conference,this paper uses a method based on conditional random field to automatically identify and obtain the signal words.Fused with cross event theory,it uses classification algorithm of maximum entropy model for temporal relation identification of the words and other linguistic features.Meanwhile,it uses the inference method of constraint propagation to solve the problem of sparse corpus.Experimental results show that the automatic recognition accuracy of signal words by using the method based on conditional random field is 69.21%,and the recognition accuracy of the temporal relation by method fused with cross event theory is 84.7%,which means that the proposed method can improve the recognition effect.
  • DONG Leigang,LIU Guohua
    Computer Engineering. 2017, 43(6): 195-201,206. https://doi.org/10.3969/j.issn.1000-3428.2017.06.031
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Existing algorithms for querying combinatorial Skyline have low pruning efficiency and unsatisfactory computing speed for low-dimensional data set.Aiming at these problems,this paper proposes a fast algorithm for solving and updating combinatorial Skyline.It defines the related concept of dominant area and presents some pruning rules to prune lots of useless data.For points added or deleted,based on existing result,it takes corresponding operations according to the data partition,so as to update query results quickly.Experimental results show that the proposed algorithm can improve pruning efficiency and computation speed effectively.
  • QIN Dongdong,CHEN Zhijun,YAN Xueqin
    Computer Engineering. 2017, 43(6): 202-206. https://doi.org/10.3969/j.issn.1000-3428.2017.06.032
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The traditional hard and soft threshold function is not continuous at the threshold,which is easy to generate oscillation,and the estimated wavelet coefficients and the actual wavelet coefficient constant deviation,there are over kill phenomenon.In view of the above shortcomings,a new threshold function is proposed,which can not only be continuous at the threshold,but also estimate the wavelet coefficients.Through adding threshold adjustment factors,according to different length sampling to adjust threshold factors,it solves the problem of large fixed threshold.Experimental results show that the improved threshold function,both in visual effect,or in the Peak Signal-to-Noise Ratio(SNR) and Mean Square Error(MSE) performance are better than other threshold functions.
  • WANG Kuikui,YU Zhenming
    Computer Engineering. 2017, 43(6): 207-211,218. https://doi.org/10.3969/j.issn.1000-3428.2017.06.033
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Aiming at the problem of present related image blur measurement method cannot detect regions with flat or clear cextured,a novel approach of regional blur identification is proposed and applied to moving object detection in static images with motion blur.Image patches are got from one image,the number of zero Discrete Cosine Transform(DCT) coefficients at every patch is counted.The high number is regard as clear region,low number is blur region.Local structure tensor of every pixel belonging to blur region got by rough detection is computed.To get local directional coherence of singular values,structure tensor is decomposed by singular value decomposition.Experimental results show that the proposed method measures the local blur more accurately than the existing method.
  • YUE Juan,LI Fanming,GE Jun,LIU Shijian
    Computer Engineering. 2017, 43(6): 212-218. https://doi.org/10.3969/j.issn.1000-3428.2017.06.034
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Current mainstream methods for mesh piecewise reconstruction cannot directly be used to reconstruct objects with sharp edges,a method for mesh piecewise reconstruction based on feature detection is proposed.It builds the object’s triangular grid model,and performs an operation of feature recognition on the grid model,and calculates the normal vector set of each grid vertex,Based on the vertexes’ positions and normal vector sets,it implements a quadratic interpolation on each triangle patch,so the 3D reconstruction of the object is achieved.Experimental results show that it can attain C0 continuity,which meets the need of visualization.Compared with that of the mainstream methods,its Signal-to-Noise Ratio(SNR) is more than 2 times bigger and its MSE is decreased one order of magnitude at least.
  • YE Run,DENG Yu
    Computer Engineering. 2017, 43(6): 219-224,229. https://doi.org/10.3969/j.issn.1000-3428.2017.06.035
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Aiming at the inherent shortcomings that many existing target tracking methods lack of stability in non controlled environment and the detecting-tracking module are separate,a multi-target tracking method based on sparseness detector and network data association technique is proposed.The moving space of the target is discretized,and the target is projected onto the image plane for each possible 3D position.The dictionary is built after formation of code.The model is extended to multiple classes.The optimal solution is achieved by assigning coupling equation to the sub problems and coordinating local solutions.The minimum cost flow problem of data association is solved by network simplex algorithm.BU-Marathon,PETS2009 and other public data sets are used in the experiments.Experimental results show that,compared with the multiple target tracking algorithms with occlusion handling,the proposed method has higher tracking accuracy and less false detection rate and missing detection rate.
  • DENG Ningyi,SHEN Zhiqiang,GUO Yuefei
    Computer Engineering. 2017, 43(6): 225-229. https://doi.org/10.3969/j.issn.1000-3428.2017.06.036
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Most existing saliency detection algorithms are based on low-level image features,therefore the results can be disturbed when applying on complicated images.In order to solve this problem,this paper proposes a saliency detection algorithm using category prior and deep neural network.Based on the feature maps generated by Deep Neural Network(DNN),it selects those parts which contribute to the classification results of pre-training,and generates the saliency map according to the preserved spatial information.With the color contrasting and image segmentation,the test images get the saliency segmentation.Experimental results on public dataset of Internet images show that compared with IT and SR algorithms,the average exact rate,call rate and F value of this algorithm have obvious improvement.
  • LU Yuanyao,GUO Zhenfang
    Computer Engineering. 2017, 43(6): 230-235. https://doi.org/10.3969/j.issn.1000-3428.2017.06.037
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    This paper proposes a lip contour extraction algorithm based on a multi-color space processing,Bacterial Foraging Optimization Algorithm(BFOA) and key endpoints localization methods.The color space conversion is applied for the lip image,color enhancement and color segmentation is used in different color space over the lip image.The method of BFOA combined with Kapur maximum entropy threshold method are applied to obtain the best threshold of the lip region,achieving accurate lip segmentation.An adaptive binarization is performed to the resulted image to extract key points along the lip edge.Using the key points as inputs,it fits the lip edge contour based on the least square criterion and extracts outer contour of the lip precisely.Experimental result shows the strong lip segmentation robustness and higher recognition accuracy of the proposed method.
  • ZHOU Junna,CHEN Wei,WANG Ke,CAI Changzheng
    Computer Engineering. 2017, 43(6): 236-240. https://doi.org/10.3969/j.issn.1000-3428.2017.06.038
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Tracking-learning-detecting(TLD) object tracking algorithm can achieve a long time online tracking.But when the appearance changes of the object by the plane rotation and the object is occluded,the TLD tracking algorithm tracks drift in the process of tracking.Aiming at above these problems,the tracking module of the TLD algorithm adopts sparse prototype,and the Spares Prototype(SP)-TLD object tracking algorithm is proposed.When the object deformation caused by the plane rotation is present,the SP-TLD object tracking algorithm accurately tracks the object by changing the coordinate position of the affine transformation and avoids tracking drift.When the object is occluded,the algorithm determines the unoccluded part and occluded part according to the Principal Component Analysis(PCA) basis vectors and trivial templates of the object and tracks the object accurately.Experimental results show that the SP-TLD algorithm has higher tracking accuracy and stronger robustness that the TLD algorthm.
  • ZHAO Tao,GUO Meng,GU Yaliu,ZHANG Yang
    Computer Engineering. 2017, 43(6): 241-246. https://doi.org/10.3969/j.issn.1000-3428.2017.06.039
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to improve the real-time performance of Data Acquisition Controller(DAC) based on general PC,a multi-channel real-time data acquisition controller is designed.By using the Field Programmable Gate Array(FPGA) hardware technology,the data acquisition of the finite state machine is used to replace the CPU serial instruction to realize channel switching and data cache,which can reduce the load of the system.Based on the new data acquisition controller architecture,a timestamp synchronization mechanism is proposed to improve the real-time performance of the DAC control system with the multi-level buffer mechanism,and the logic of the controller is successfully transplanted to the Cyclone IV chip for implemention.The test results show that the proposed controller can alleviate the load of CPU and operating system,and improve the real-time performance of data and the precision of sampling bandwidth.
  • LIU Zhao,SONG Libin,GENG Meixiao,YU Tao,WANG Zengxi,GUO Kai
    Computer Engineering. 2017, 43(6): 247-252,258. https://doi.org/10.3969/j.issn.1000-3428.2017.06.040
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    To make the dancing robot plan a dynamic path based on the motion of pedestrians and enhance its ability of getting along with people,an indoor pedestrian tracking method based on laser radar is proposed.The source data collected by the laser is gotten and preprocessed.According to the distribution characteristics of laser data,the DBSCAN algorithm is optimized to cluster the data rapidly to realize the segmentation of the environmental information.A pedestrian recognition method is proposed based on the distance from cluster to laser radar and the pedestrian width.The position of the pedestrian cluster is used as the initial tracking position.The laser data is displayed graphically and converted into video data.The particle filtering algorithm is applied to track the pedestrian,and the trajectory is drawn at the same time.Experimental results show that the proposed method has good effect on pedestrian recognition and tracking.It also has a good real-time performance.
  • ZHANG Xixiang,ZHAO Huan
    Computer Engineering. 2017, 43(6): 253-258. https://doi.org/10.3969/j.issn.1000-3428.2017.06.041
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to improve the prediction accuracy of speech personality,this paper proposes a method of speech personality prediction based on random forest model.The candidate prosodic feature set for speech personality prediction is selected,and the prosodic feature set is sampled by bootstrap method.According to the Gini coefficient,the optimal prosodic feature set is selected for each decision tree node,and the random forest model corresponding to each dimension personality is constructed to achieve speech personality prediction.The results of experiment on open source speech personality corpus show that this method has better performance in terms of prediction accuracy than other speech personality prediction methods.
  • XU Yaosong,LI Meng
    Computer Engineering. 2017, 43(6): 259-263. https://doi.org/10.3969/j.issn.1000-3428.2017.06.042
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to solve the problem that the traditional generalized cross correlation time delay estimation algorithm has poor anti-noise ability and low precision of time delay estimation in complex environment with reverberation and noise,a time delay estimation algorithm based on Forward Spatial Linear Prediction(FSLP) is proposed.Combined with the redundant information and the spatial correlation matrix,the estimation method of the time difference of arrival time in adverse acoustic environment is improved.The space-time prediction algorithm which is based on the spatial and temporal correlation of the speech signal source in the microphone array output can suppress the noise and minimize speech distortion.Simulation results show that,compared with the traditional generalized cross correlation time delay estimation algorithm,the time delay estimation error rate of the proposed algorithm is lower and the stability is higher.
  • QIAO Zhi,XU Xiangyu,YU Jiadi,LI Minglu
    Computer Engineering. 2017, 43(6): 264-269,275. https://doi.org/10.3969/j.issn.1000-3428.2017.06.043
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to implement the fine-grained detection and identification of inattentive driving,this paper proposes an inattentive driving detection system based on intelligent terminal,which does real-time detection for inattentive driving behavior by leveraging speaker and microphone of smartphone.According to the definition of inattentive driving behavior,it abstracts four specific actions:turn around,take back,pick up and hand continuous small movement,and analyzes the unique patterns of the Doppler effects from these actions.It extracts data features by using Principal Component Analysis(PCA)method,uses Support Vector Machine(SVM)to train the features and outputs a classifier model which conducts fine-grained identification.Experimental results demonstrate that the proposed system can achieve accuracy of 93.72% in average when judging inattentive driving,which has better detection performance.
  • YIN Zhonghai,ZHAO Xin,ZHOU Cheng,CHU Yanan,WANG Lei
    Computer Engineering. 2017, 43(6): 270-275. https://doi.org/10.3969/j.issn.1000-3428.2017.06.044
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The state diagram is unable to clearly describe the interaction between components during state transition in Cyber Physical System,resulting in low system efficiency.In order to solve this problem,a hierarchical representation of state transition is proposed.State transition is divided into component affected layer and state transition layer.In component affected layer,directed graph is used to describe the relationship between components.State transition layer is used to describe the transition,and the experiment of automobile cooperative adaptive cruise control is performed.The experimental results show that,compared with the state chart method,the proposed method adjusts the acceleration speed adaptively according to the distance between the two vehicles and the speed difference,enabling the car to enter the cruise state faster.
  • LIANG Yuan,CHE Ming
    Computer Engineering. 2017, 43(6): 276-280,288. https://doi.org/10.3969/j.issn.1000-3428.2017.06.045
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The traditional fuzzy equivalence relation clustering method cannot clusteraccording tospecific constraints,so that the clustering results have low accuracy,anddonot meet the requirement.In order to solve this problem,based on traditional fuzzy equivalence relation clustering method,by means of distance constraints for preprocessing and expanding the data set,this paper proposes a novel fuzzy clustering method.According to the Euclid distance and constraint conditions,the inter-data relationship for each data is established,so as to describe whether the distance between data points meet the constraint conditions.These relationships are regarded as the additional dimension.The original data set is updated and the similarity equation is reconstructed to obtain the corresponding similarity matrix and cluster based on fuzzy equivalence relation.Experimental results on real dataset show that compared with the traditional unsupervised fuzzy equivalence relation clustering method,the proposed method overcomes the defectthat it is unable to accurately clusteraccording tospecific constraints,and the clustering results can bettersatisfy the constraint conditionsand they have higher accuracy.
  • LI Zhi,JIAO Qinqin,ZHOU Yufeng
    Computer Engineering. 2017, 43(6): 281-288. https://doi.org/10.3969/j.issn.1000-3428.2017.06.046
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The research on Location-Allocation Problem(LAP) of post-earthquake emergency facilities helps to improve the efficiency of disaster relief.Taking maximization of demand utilities and fairness of material allocation as goals,a multi-objective LAP model of post-earthquake emergency facilities is built in this paper based on a mixed integer programming method.Considering characteristics of the model,a non-dominated sorting multi-objective genetic algorithm based on matrix encoding and niching technique is designed to solve the problem.Numerical results show that the algorithm can give the Pareto front.Decision makers can balance multiple objectives according to their preferences and actual needs,and select the appropriate decision making scheme on the Pareto front.
  • WANG Zhiyuan,CHENG Lan,XIE Gang
    Computer Engineering. 2017, 43(6): 289-295. https://doi.org/10.3969/j.issn.1000-3428.2017.06.047
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Aiming at the problem of particle depletion in traditional Particle Filtering(PF),a PF algorithm based on Adaptive Differential Evolution(ADE) is proposed.The ADE algorithm instead of the re-sampling strategy is used to generate new particles in PF,which promotes the particles moving toward the region with high likelihood in the state posterior probability density function,and increases the diversity of the particles.A nonlinear adaptive control strategy is adopted to adjust the mutation factor and the crossover factor for improving the ability of optimization of DE in PF.Simulation results show that applied for multipath estimation,the proposed algorithm can overcome the problem of particle depletion.Compared with algorithms of PF,Extended Kalman Filtering(EKF) and Differential Evolution Particle Filtering(DE-FP),the proposed algorithm has better multipath estimation performance.
  • JI Yimu,CHAI Bozhou,YANG Luokun,ZHU Tonghui,LI Wenfeng
    Computer Engineering. 2017, 43(6): 296-300. https://doi.org/10.3969/j.issn.1000-3428.2017.06.048
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The traditional rail transit cluster dispatching service system has low peak communication rate and cannot realize the high-frequency voice communication and basic video communication in the daily dispatching.Aiming at this problem,this paper designs a new dispatching system which uses the technology of 4G Time Division Long Term Evolution(TD-LTE),optimized mechanism of database connection pool and the strategy of I/O Completion Port(IOCP).Simulation results show that this system can effectively reduce communication delay and improve concurrency.It can meet the dispatching needs of large capacity,high rate and multimedia demand of modern rail cluster communication network.
  • HUANG Zhezhi,WANG An,ZENG Qinghua,LONG Zhiqiang
    Computer Engineering. 2017, 43(6): 301-307. https://doi.org/10.3969/j.issn.1000-3428.2017.06.049
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The fault diagnosis method based on modeling is intuitional,reconfigurable and also effective.However,modeling the signal/function of electrical system can hardly locate the physical position of the fault.Therefore,this paper expands Logical Physical Synthetically(LPS) model to the application of fault diagnosis in general electrical system.It proposes the principle of LPS modeling in electric system,and presents a hierarchical modeling,merge processing in Replaceable Unit(RU),signal feedback closed-loop path,automatic fault mode correlation and other modeling processing strategies.Taking a triangle wave generator circuit as an example,this paper also shows the modeling process of LPS modeling in electrical system,which indicates that LPS modeling method can be applied in general fault diagnosis in electrical system.The diagnostic process and the result can be visually displayed.
  • ZHANG Cheng,QI Wangdong,WEI Li,ZHAO Yuexin
    Computer Engineering. 2017, 43(6): 308-312. https://doi.org/10.3969/j.issn.1000-3428.2017.06.050
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to improve the practicality of Radio Interferometric Ranging(RIR) in severe multipath environments,a model of multipath-induced error in range estimation is established and the influence of system parameters on multipath errors is investigated accordingly.It is revealed that there is a threshold effect in the relationship between multipath errors and system parameters.The ranging performance of RIR is evaluated with universal software radio peripheral N210 software radio platform in two different indoor scenarios.Experimental results indicate that the root mean square errors of RIR are reduced to about 1 meter as measurement bandwidth hits 160 MHz and the number of frequencies arrives at 40,verifying the feasibility and potential of RIR in indoor environments.
  • SONG Qingzeng, Lü Huayang,ZHAO Lei,WANG Jiangfeng
    Computer Engineering. 2017, 43(6): 313-321. https://doi.org/10.3969/j.issn.1000-3428.2017.06.051
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Accurately measuring and analyzing the power consumption characteristics of the Xeon Phi coprocessor is a basic prerequisite of power management and optimization.However,accurately profiling and analyzing the power consumption of parallel programs running on Xeon Phi is not straight forward.This paper uses special measurement equipment to measure power consumption of Xeon Phi coprocessor.The real-time voltage and current of 14 power supply channels are completely measured,which can get the power consumption of coprocessor by calculation.Based on the measuared data some power characteristic about starting-up power,idle power,thread power and memory hierarchy power.Experimental results show that this paper provides credible power consumption data for power optimization,it can guide the power optimization of processors Xeon Phi.