Loading...
  • 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

Current Issue

    • 论文
      Three grid job scheduling methods heuristics
      for data-intensive and computing-intensive jobs            
      HAO Yongsheng 1,LU Junwen2,LIU Guanfeng3,WEN Na4
      2014, 36(08): 1423-1429. doi:
      Abstract ( 121 )   PDF (1305KB) ( 234 )     

      Most of the existing grid job scheduling methods focus on either dataintensive jobs or computingintensive jobs.In the dynamic environment where every job has its deadline,we extend the traditional grid job scheduling methods to propose three new grid job scheduling methods:Eminmin,Ebest and Esufferage.The three methods are validated on the Grid model with clusters that are connected by the high speed network.Simulation results demonstrate that our proposed methods are better than Minmin.The comparison between the three methods and ASJS shows that,Eminmin reduces the waiting time and makespan,Esufferage reduces the waiting time and the makespan greatly with the sacrifice of some jobs,and Ebest gives the same performance in unfinished jobs but has a larger value in waiting time and makespan than ASJS. In general,Eminmin has a better performance than Minmin and ASJS.

      Simulation research of monitoring system
      reliability based on cloud model         
      SHAO Ying1,YUAN Lijun2,ZHUANG Biao3
      2014, 36(08): 1430-1434. doi:
      Abstract ( 100 )   PDF (627KB) ( 234 )     

      There has not been an effective simulation and analysis method for predicting the reliability of monitoring system for a long time.The cloud model theory and the fault tree analysis (FTA) are used to transform elements into environmental factors that affect the reliability of the monitoring system.These elements such as temperature,humidity,voltage and stress of parts can be gathered to meet the entire monitoring system cloud model in order to obtain a new method to predict the reliability of the monitoring system.Through the test of an actual monitoring system, the effectiveness of the simulation and analysis method is fully verified.According to the reliability objective of the monitoring system,the constructor is improved to obtain better reliability effect.

      Accelerating 3D GVF field computation on #br# Xeon Phi using stencil optimization         
      QI Jin, LI Kuan, YANG Canqun, DU Yunfei
      2014, 36(08): 1435-1440. doi:
      Abstract ( 122 )   PDF (843KB) ( 166 )     

      3D Gradient Vector Flow (GVF) field has wide applications in many image processing algorithms. The computation of GVF field typically needs several iterations and is rather time consuming. Therefore, it is important and meaningful to improve the computation speed of 3D GVF field. The data level parallelism and thread level parallelism are introduced to accelerate the GVF field computation procedure on Intel Xeon Phi many core integrated platform for the first time. Meanwhile, GVF field computation is a kind of stencil computation, whose computationmemory access ratio is low. A novel cache blocking strategy is proposed to fully utilize the L2 cache of Xeon Phi architecture,and to improve the computation speed of GVF field. The experimental results show that the proposed optimizations could effectively improve the speed of GVF filed computation. Especially, for a 5123 3D image, compared with the performance obtained by a 2.6G Hz 8 core 16threads Intel Xeon E52670 CPU, the speedup achieved on Xeon Phi is 2.77X.

      论文
      An isomorphic overcomplete dictionary learning
      method for superresolution reconstruction       
      XIE Baoling1,XU Guoming1,2
      2014, 36(08): 1441-1446. doi:
      Abstract ( 126 )   PDF (803KB) ( 224 )     

      Constructing an appropriate overcomplete dictionary is one of the key problems of superresolution based on sparse representation. In the maximum likelihood estimation principle, an isomorphic overcomplete dictionary learning model based on mixture Gaussian is proposed. Firstly, the sparse coding residual of the model is described by the weight l2norm and the weight matrix is designed by the residual. Secondly,the isomorphic coupled dictionary learning problem is translated into the single dictionary learning problem. The dictionary is learned by the alternate and iterative strategy using sparse coding and dictionary updating. An interiorpoint method is used in sparse coding and Lagrange dual is used in dictionary updating. Finally, the learned dictionary is used in the superresolution experiment,and compared with other methods.The experimental results demonstrate the effectiveness of the proposed method.

      An efficient scalable metadata
      management method in cloud computing           
      HUANG Bin1,2,PENG Yuxing3
      2014, 36(08): 1447-1454. doi:
      Abstract ( 105 )   PDF (952KB) ( 164 )     

      Aiming at the lower performance problem of the existing metadata management methods in cloud computing,an efficient scalable metadata management method is proposed in cloud computing.Firstly,a dynamic binary mapping tree is used to achieve the precise positioning of the metadata server.Secondly,a lazy update technique is adopted to dynamically update the binary mapping tree.Finally,a dynamic K tree is proposed to improve the speed of selecting migrated metadata during MDS splitting.The experimental results show that the method can effectively improve the efficiency of the scalable metadata management method in cloud computing.

      Design and optimization of 14Gbps
      high-speed channel based on simulation           
      SUN Yan,LI Tiejun,AI Mingzhe,HU Jun,CAO Yuesheng
      2014, 36(08): 1455-1461. doi:
      Abstract ( 169 )   PDF (1721KB) ( 165 )     

      The signal transmission rate is an important index of high performance computing. As signal transmission rate of modern high performance computing systems is continuously increasing, which reached to, even over 10Gbps, the design of highspeed channel faces severe challenge with the rapidly increased signal rate. Based on the guidance of the simulation analysis on signal integrity, a 14Gbps high-speed channel is optimized and designed. Through combining the manual 3D model with the extracted PCB model, the simulation analysis is speeded up. The full channel co-simulation is used to forecast the overall performance and the bottleneck of the high-speed channel, and we focus on optimizing the hole, dielectric materials, line width and line space etc. Finally, we successfully carry out the stable transmission of 14Gbps high-speed signals.

      Online updating of self-adaptive middleware
      based on reinforcement learning       
      WANG Jianjun,LIU Yulin
      2014, 36(08): 1462-1468. doi:
      Abstract ( 95 )   PDF (663KB) ( 244 )     

      One common approach of selfadaptive middleware is to incorporate a control loop that monitors, analyzes, decides and executes over a target system with predefined strategies. Such approach is an offline adaptation where strategies or adaptive models are statically determined so as not to change with environment. Aiming at the problem, an online updating mechanism of selfadaptive middleware based on reinforcement learning is proposed to solve the problems of conflict resolution and realtime system effectiveness evaluation, and an online updating method of selfadaptive policy based on reinforcement learning is designed, thus enhancing intelligence, flexibility and reaction capability. Finally, the corresponding system OUSAM is implemented and the effectiveness and feasibility of the mechanism is validated on OUSAM.

      Intrusion detection based on the adaptive
      evolutionary neural network algorithm                   
      YANG Hongyu,ZHAO Mingrui,XIE Lixia
      2014, 36(08): 1469-1475. doi:
      Abstract ( 119 )   PDF (606KB) ( 203 )     

      Aiming at the problem of low detection rate existed in current intrusion detection systems,an adaptive evolutionary neural network algorithm (AENNA) based on the genetic algorithm and the BP algorithm is proposed.Firstly,considering the characteristic of probabilistic jumping property and local search ability in the simulated annealing algorithm,the genetic algorithm is improved.Secondly, using the dual population evolution rule,the algorithm optimizes the weight and the network structure of the BP neural network simultaneously.An adaptive neural network training method is designed through improving the crossover operator and mutation operator of the genetic algorithm.Experimental results show that the AENNA based intrusion detection method can effectively improve the detection rate and reduce the false positive rate.

      Design of taxi dispatch system based on Internet of Things           
      YAO Zhongmin,YAO Tianjian,LONG Zhaopeng,LI Qiang
      2014, 36(08): 1476-1481. doi:
      Abstract ( 109 )   PDF (1072KB) ( 203 )     

      The current taxi dispatch system based on taxi stop board has the shortages that passengers jump the queue frequently, the scheduling result is not optimized and appointments are broken by both passengers and taxi drivers. In order to solve these problems, a taxi dispatch system based on Internet of Things (IoT) technology is proposed. The system adds Radio Frequency Identification (RFID) reader as the perception layer, General Packet Radio Service (GPRS) gateway modules as the transport layer and scheduling management center as the application layer, etc. The test results show that, each taxi stop of the system distributed in the city can schedule the nearby taxis. The dispatch management center can realize the management and query of scheduling results of each taxi stop.

      Analysis of the worm propagation model with constant immigration           
      HU Mingsheng1,2,JIA Suimin1,CHEN Qiaoling1,JIA Zhijuan3,HONG liu2
      2014, 36(08): 1482-1485. doi:
      Abstract ( 113 )   PDF (655KB) ( 185 )     

      According to worm propagation,a SIRS propagation model of susceptible hosts with constant recruitment and standard incidence rate is proposed. Considering that the total host count is dynamically changed during the propagation, the qualitative and stability theory of the differential equation is applied to analyze the model. Not only the propagation regularity of worm virus but also the effects of different factors on the worm virus control are discussed. And the Abilene network is used to analyze the effects of network topology on worm propagation rate. At last, the worm data provided by CAIDA are used to test the model.

      Research of the linkage system of firewall
      and intrusion detection system based on ACE and SSL                   
      MA Zhanfei1,YIN Chuanzhuo2
      2014, 36(08): 1486-1492. doi:
      Abstract ( 100 )   PDF (611KB) ( 147 )     

      With the rapid development of Internet, some intelligent attack methods and techniques are increasing. The network is easily attacked by hackers or malicious software. The safty problem is increasingly outstanding in computer network. The traditional technologies of firewalls and Intrusion Detection Systems (IDSs) own poor security, high false alarm rate, and low level of intelligence. Considering the demands of obtaining integrity and dynamics in network security, a novel linkage system model of firewall and IDS based on open communication platform of ACE (Adaptive Communication Environment) and SSL (Secure Socket Layer) is proposed. This system model combines the advantages of firewall and IDS, and uses the encrypted information transmission mechanism, and the policy management mechanism, and the associated linkage analysis algorithms to ensure the reliability, integrity and confidentiality of the transmitted information. Experimental results show that the linkage system can effectively prevent network from attacks, and possesses better cooperativeness, universalness and expansibility.:

      An improved scheme of end-to-end
      TCP congestion control in Ad Hoc network         
      TENG Yanping,WANG Haizhen,JIN Mei,LI Dahui
      2014, 36(08): 1493-1499. doi:
      Abstract ( 165 )   PDF (950KB) ( 220 )     

      The traditional TCP protocol is designed for wired networks, and it is assumed that the packet loss is caused by network congestion. However, in Ad Hoc network, in addition to the congestion loss, other noncongestion factors can also cause packet loss. The main factors affecting TCP performance of Ad Hoc Network are analyzed, and an endtoend TCP congestion control scheme (ImpMMJI) is proposed, which is based on the original congestion control scheme of MMJI. According to the forward path hops, the scheme adaptively adjusts the congestion window size and prevents the excessive growth of the congestion window. When the route changes or link interruption occurs, the scheme can calculate cwnd and ssthresh values of the congestion window in order to ensure the consistency of the TCP connection load rate when the routing is reconstructed. Besides, the scheme adds the state sign bit in TCP header of the ACK reply packet, which combines multiple measurement parameters to judge the network state so as to improve the network state recognition accuracy and make the sender respond in real time. The simulation results show that the scheme can improve the network throughput and the TCP performance obviously.

      Design and implementation of
      an interactive system based on Bluetooth            
      WANG Zhaoshun1,XU Baiquan2
      2014, 36(08): 1500-1504. doi:
      Abstract ( 128 )   PDF (676KB) ( 178 )     

      With the development of the wireless connectivity technology and the mobile terminal technologies, mobile terminal users such as mobile phone users ask Bluetooth products for higher requirements. Bluetooth uses the push technology to transfer data, and users only have authority to choose whether to receive the data but cannot browse the data from the Bluetooth transmitter independently and interactively. An interactive system based on the J2ME and Bluetooth technologies is introduced. It utilizes the advantage of crossplatform of J2ME as well as the advantages of easy of use and practicability of Bluetooth technology to transfer data. Based on the JSR82 Java Bluetooth wireless technology API provided in the specification, the design scheme and system structure of the proposed Bluetooth interactive system is given. Based on the J2ME wireless application development kit and the MIDlet programming model, the interactive system is implemented.

      Research on extracting text from MS-DOC files          
      HUANG Bugen1,FU Juan2
      2014, 36(08): 1505-1511. doi:
      Abstract ( 108 )   PDF (1023KB) ( 217 )     

      Keyword search is widely used in intelligence analysis, search engine and computer forensics. However, sometimes searching key words in MSDOC files may fail to find out some matches, which are usually called false negatives. Microsoft compound document is composed by a series of stream stored in sectors. The streams and the sectors are managed through the directory and the sector allocation table. The text is stored in the MSDOC file WordDocument stream, text storage is not necessarily continuous, and the Table stream records the block information. Keyword may be stored in different sectors that are not adjacent. Even the sectors are adjacent, the part of the keyword may be compressed, but the other part is not compressed. These cause the false negatives. Firstly, texts are extracted from the WordDocument  stream based on the block information in the Table stream, and they are encoded uniformly. Secondly, a keyword search is carried out. These two steps can avoid the false negative.

      The fault diagnosis of the wireless sequence
      control system based on fuzzy Petri nets         
      XU Chengyan,BAI Yan,WANG Renshu
      2014, 36(08): 1512-1518. doi:
      Abstract ( 99 )   PDF (1016KB) ( 172 )     

      In order to diagnose faults in the wireless sequence control system which is based on the wireless sensor network and the traditional sequence control, the system structure is analyzed and its faults transmission rules are proposed. A fuzzy reasoning mechanism with an improvement of adding threshold judgment is proposed to make up the weakness of the confidence degree method of fuzzy Petri net in faults diagnosis. The improved fuzzy petri net diagnosis method is applied on the wireless sequence control system to diagnose the controller’s fault occurrence probability. It is concluded that the control logic has the largest fault occurrence probability, which is in line with the field test result.Key words:

      Eliminating redundant arguments
      of RED by feature analysis       
      WU Bin,MA Jitao,WU Ping,TAN Peng
      2014, 36(08): 1519-1523. doi:
      Abstract ( 95 )   PDF (749KB) ( 130 )     

      The RED algorithm is the basis of network congestion control.By using this algorithm, more than one parameter,such as queue average length,limit value of dropping packets,average size of packets,need to be loaded. What’s more,arguments’values are set without clear rules and the theoretical basis and unreasonable values weak the effect of the congestion control algorithm. The relationship of two system performance indicators of arrival rate of packets and the capability of data processing are analyzed by the way of feature analysis and build up a map relation between indicators and arguments of RED.Furthermore,excess parameters are combined and the ranges of values are confirmed by using the linear relation among parameters.The congestion control effect of the algorithm is improved in the network environment and increase the practicality of RED.

      The analysis of a SIQR epidemic model
      in scale-free complex networks           
      YE Zhiyong,LIU Yuan,ZHAO Yanyong
      2014, 36(08): 1524-1527. doi:
      Abstract ( 115 )   PDF (470KB) ( 148 )     

      A SIQR epidemic model with quarantine in scalefree complex networks is investigated.by means of the mean-field theory,the spread of the disease is studied.Threshold R0 is obtained by calculation,and it is proved that the disappearance or eruption of the disease is determined by threshold R0. .Computer simulation indicates that reducing infection rate and improving the isolation rate of infected nodes can effectively control the spread of infectious diseases.

      Research of the geological data nonlinear analysis problem based
      on the SPDS algorithm in the multilayer feed-forward network             
      DAI Ke,ZHANG Shaozhong,JIANG Bo,BAI Ying,WANG Xiaomei
      2014, 36(08): 1528-1532. doi:
      Abstract ( 75 )   PDF (470KB) ( 154 )     

      Abstract:The multilayer feedforward neural network is good at solving the nonlinear analysis problem, but for complex geological data, the common training algorithm is difficult to converge. The SPDS algorithm is proposed to solve the problem of the geological data nonlinear analysis. Simulation results show that the proposed algorithm can be better to solve the problem in the multilayer feedforward neural network.

      Study on homologous relationship in the
      precomplete sets of partial multiple-valued logic        
      GONG Zhiwei
      2014, 36(08): 1533-1537. doi:
      Abstract ( 103 )   PDF (360KB) ( 151 )     

      By studying the inner link amongst precomplete sets, a homologous concept between two mary relationship is proposed firstly. Secondly, the property of the homologous relationship among precomplete sets is analyzed and proved. Finally, the effect of this property in the proof of the minimal covering of precomplete sets is introduced. 

      Research on visual ergonomics evaluation of driller control room          
      DENG Li1,2,YU Suihuai1,CHU Jianjie1,ZHANG Wanyu1
      2014, 36(08): 1538-1543. doi:
      Abstract ( 95 )   PDF (1427KB) ( 153 )     

      To solve the problem of artificial calculation and mapping error in traditional visual ergonomics evaluation, based on the analysis of the human vision characteristics, the eyellipse and H point used in the car vision design principle are introduced, and the visual ergonomics evaluation model of the oil rig driller control room is established by combining the computer aided ergonomics design technology. Taking CATIA as the secondary development platform, the visual ergonomics evaluation system of the driller control room is programmed by VB, which realizes the accurate and rapid positioning of eyes with different percentile human body model. Finally, a case illustrates that the system is practical and able to satisfy the visual assessment requirements of the humanmachine interface layout of the driller control room.

      论文
      论文
      论文