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
      Research on FPGAbased acceleration of
      finite difference time domain algorithms            
      SONG Qingzeng1,ZHANG Jinzhu2,WU Jigang1
      2013, 35(9): 1-6. doi:
      Abstract ( 183 )   PDF (739KB) ( 335 )     

      As realtime predicting electromagnetic behaviors are required more and more in some realtime embedded systems, Finite Difference Time Domain (FDTD) is implemented in FPGA hardware to increase the computational speed. A novel method to implement FDTD on FPGA is proposed. The filters technology is used to rewrite the FDTD algorithm. In this way, the solution of the FDTD is transformed to the design of the corresponding hardware filters. Therefore, a suitable filter can be designed to implement the FDTD method. Experimental results show that, compared with the software implementation, the hardware implementation of FDTD obtains about 5 times speedup and can taking full advantages of FPGA computing performance. This study can further widen the application field of FDTD, and especially extend it to the fields that are not applied before due to the computing ability.

      A data service supporting variable input and output parameters        
      ZHANG Feng1,2,CHEN Xin1,WEI Yongshan1
      2013, 35(9): 7-14. doi:
      Abstract ( 127 )   PDF (497KB) ( 229 )     

      Traditional Web services provide operations with fixed input and output parameters, thus they cannot meet various queries based on the content of data resources. A data service supporting variable input and output parameters is proposed, and various queries based on the content of the data can be met. Firstly, description of the contents of data supplied by the service is inserted into the service description, and data provided by the service can be determined by the description. Secondly, the operation supporting variable input and output parameters is defined. The input of this operation is a query containing input parameters, output parameters and constraints, and the output of this operation is the data corresponding to output parameters in the query. Finally, the method for servitization of relational data resources is given, and the effectiveness of the method is validated through experiments.

      Parallel algorithm for PAML on multicore platform        
      YANG Ju,WU Zhuofeng,WANG Gang,LIU Xiaoguang
      2013, 35(9): 15-19. doi:
      Abstract ( 171 )   PDF (489KB) ( 226 )     

      PAML, which is widely used, is a package of programs for maximum likelihood phylogenetic analysis of protein and DNA sequences. However, its calculation is time consuming because of too many parameters and complex models. The paper focus on the most important part in PAML: codeml. After analyzing the algorithm and finding program bottleneck by profiling, we parallelize the bottlenecks based on the capacity of modern multicore CPU and SIMD technology in order to reduce the overall operating time. The experimental results on both real and simulated data show that our parallel algorithm for PAML can improve the calculation efficiency. The final speedup is up to 7.94.

      Research of MPI nonblocking broadcast
      algorithm and its performance analysis     
      YAN Xinkai,HAO Ziyu,WU Dong,XIE Xianghui
      2013, 35(9): 20-26. doi:
      Abstract ( 129 )   PDF (1691KB) ( 275 )     

      The MPI3.0 standard contains nonblocking collectives. Nonblocking collectives captures the features of both nonblocking communication and collective communication. Compare with blocking collectives, nonblocking collectives have lower synchronization overhead and can realize the overlap between computing and communication in the program so as to improve the performance. The paper describes different algorithm implementations of broadcast, compares the lowlevel management between nonblocking and blocking broadcast, analyzes the experiment results and finally offers some optimizations.

      Research on operating system scalability for search service         
      ZHAO Yulong,WANG Lei,WANG Huan
      2013, 35(9): 27-35. doi:
      Abstract ( 110 )   PDF (1041KB) ( 227 )     

      The scalability of the operating system is the ability that the performance of operating system grows with the increase of the number of CPU cores. Analyzing current mature operating systems to find the performance bottleneck in multicore environments and eliminate it has become one of the research spots of modern operating systems. Through pressure experiments of search service, it's found that the scalability of 64 cores Linux server running Nginx reverse proxy is poor and the throughput of 64 cores only increases by 3 times in comparison to 1 core. The bottleneck is that all NIC interrupts and soft interrupt handling concentrate in one core of the server. Using the NIC Flow Direction mechanisms and the Linux kernel interrupt affinity, the NIC interrupt load balancing approach is achieved. The average throughput increases by 33%. NIC interrupts load balancing in multicore environments cannot guarantee the TCP connection affinity. By modifying the code of the Linux kernel, the nonaffinity problem is solved, and the throughput keeps increasing by 9%.

      Contextaware intelligent service system for mine industry      
      XUE Xiao1,CHANG Jingkun1,ZENG Zhifeng2,AN Jiyu1
      2013, 35(9): 36-44. doi:
      Abstract ( 116 )   PDF (831KB) ( 299 )     

      With the rapid development of information technology, many information systems have been implemented in the mine industry in recent decades. However, the usage of current information services is limited, which are only accessible to managers for making decision. The majority of miners can hardly access these information services timely in order to prevent hidden risks. To address this challenge, a framework for building the contextaware intelligent service system for coal mine industry is proposed. This system includes four major components: context information collecting, context information reasoning, service configuration and invocation. For the sake of meeting the demands of the end miners, the proposed system takes full advantage of the legacy information systems, realizes more comprehensive interoperability and provides more intelligent business service & composition. In the end, the process of building service systems and the effect of implementing contextaware services are elaborated through testing the system in a simulated environment and various potential problems are also discussed.

      Deterministic regular grammars based on quantum logic      
      WANG Yongbing1,ZHANG Lixia1,LEI Hongxuan2
      2013, 35(9): 45-50. doi:
      Abstract ( 110 )   PDF (417KB) ( 206 )     

      Automata theory based on quantum logic is an important aspect of the quantum computing models. The paper introduces an orthomodular latticevalued deterministic regular grammar, and shows the equivalence between orthomodular latticevalued deterministic regular grammar and orthomodular latticevalued deterministic finite automata. Based on the equivalence, some algebraic level characterizations of orthomodular latticevalued deterministic regular languages are established and the closed properties of these orthomodular latticevalued languages in details under standard operative conditions are obtained.

      Research on dynamic optimal composition
      of contextaware logistics Web service       
      LIU Zhizhong,SONG Cheng,XUE Xiao,XU Xiuzhen
      2013, 35(9): 51-56. doi:
      Abstract ( 102 )   PDF (499KB) ( 275 )     

      The fourth party logistics services are able to combine excellent logistics services through largescale integration of resources, and provide the best logistics services scheme for the user through combination of technology, warehousing, transportation service, thus realizing the logistics service with high efficiency, low cost and realtime operation. However, how to dynamically construct reliable composite logistics service that can satisfy users’ quality requirements is the critical issue for the fourth party logistics service provider to solve. To address this issue, the paper proposes a dynamic composition method of contextaware logistics Web service. Firstly, this method decomposes the global QoS constraints into local QoS constraints based on the improved genetic algorithm, and transforms the global optimization problem into a local optimal service selection problem. Secondly, during the execution of the service process, the best logistic Web service that is available and can satisfy the local QoS constraints is selected. Finally, the effectiveness and feasibility of the method are validated by experiments.

      Scheduling model for service resource
      optimization in Internet of Things
      XU Jiuyun,WU Tongxin,NING Xiao,XING Kunming
      2013, 35(9): 57-63. doi:
      Abstract ( 115 )   PDF (593KB) ( 243 )     

      With the repaid development of the Internet of Things (IoT) and ServiceOriented Architecture (SOA), the problem of service resource competition becomes increasingly severe. In order to solve the problem, this paper studies on it from the perspective of resource optimization. Firstly, a service scheduling model is established based on the theory of resource scheduling and service selection mechanism. Secondly, the delay performance of the scheduling model is analyzed by using the queue theory. The simulation experiments demonstrate that the proposed method is efficient. It can not only guarantee the fair attainability of requesters, but also relieve the delay and delay jitter performance. Meanwhile, to some extent, it can meet the individual requirements in the case of limited service resources.

      ESSF-EGov:A holistic and interdisciplinary
      framework for e-government public service systems        
      CHEN Jian,YUAN Liu,YAN Yafei
      2013, 35(9): 64-72. doi:
      Abstract ( 102 )   PDF (1230KB) ( 189 )     

      An e-government public service system provides convenient and effective services to citizens and enterprises, through process optimization and innovations. The construction, operation and continuous improvement of e-government public service systems is a complex engineering task, involving multiple disciplines such as government decision making, IT governance, software system development and maintenance, data and knowledge management. A system framework, called ESSFEGov, with features of service systems, is described based on analysis of egovernment public service systems as well as analysis of research in related areas. The framework, based on the Zachman framework approach, is holistic in the sense that it covers the entire life cycle of an egovernment public service system, as well as links multiple perspectives of various stakeholders. The framework establishes a foundation for further developing a systematic and engineering methodology for the development of e-government public service systems. A real world application is used to demonstrate the effect of the framework for the development of a government public service portal.

      Topological study on fast detecting
      the coverage of mobile sensor network       
      HONG Feng,LIU Xu,YI Dongyun
      2013, 35(9): 73-77. doi:
      Abstract ( 115 )   PDF (503KB) ( 226 )     

      The goal of wireless sensor network is to obtain the useful information, but the quantity of the useful information is up to the coverage of the monitor area. One of the most important problems is the algorithm of the coverage. Most of existing algorithms are based on the assumption that the sensor can provide the location and relative orientation. Proposes an algorithm based on the algebra homology for detecting coverage, it can detect the condition of the coverage quickly only on the topology of the network, and there is no need to know the accurate location of the sensor. Hence, the complexity of sensor design can be reduced. Finally, the simulation results show that the algorithm can efficiently and quickly detect the blind zone of the wireless sensor network.

      Design and implementation of
      messaging service for mobile platform        
      SHAO Xiaozhe1,WU Guoquan2,WEI Jun2,3,HUANG Tao2,3
      2013, 35(9): 78-82. doi:
      Abstract ( 127 )   PDF (1369KB) ( 203 )     

      Due to the limited network connectivity and tight power constraints, it is highcost for application developers to construct and maintain a reliable and lowenergy messaging service. What we do for the mobile platform provides instant and reliable messaging service to the application developers. Moreover, owing to the feature that mobile devices own plentiful contextaware information, messaging service provides messaging based on contextaware. Our experiments demonstrate the feasibility and the low energy consumption of the messaging service.

      An empirical study of correlation between
      Web service QoS and user location         
      LIN Xiangyun1,2,LIU Xiaoqing3,TANG Mingdong1,2,CAO Buqing1,2,LIU Jianxun1,2
      2013, 35(9): 83-88. doi:
      Abstract ( 113 )   PDF (732KB) ( 198 )     

      How to accurately predict Web service QoS (Quality of Service) has recently become a hot research topic. Considering the influence of user locations on Web service QoS, some studies have proposed a couple of locationaware Web service QoS prediction methods. However, the relationship between Web service QoS and user locations is rarely empirically investigated. Therefore, an empirical study for evaluating the relationship between them is carried out, and a set of experiments on a real dataset of Web service QoS is constructed. Experimental results show that two users that are close in location are likely to observe similar QoS values on the same set of Web services, and vice versa. These findings confirm the strong correlation between Web service QoS and user locations, which can provide important evidences for the effectiveness of locationaware Web service QoS prediction.

      Impossible differential cryptanalysis
      of lightweight block cipher ESF          
      LIU Xuan,LIU Feng,MENG Shuai
      2013, 35(9): 89-93. doi:
      Abstract ( 149 )   PDF (716KB) ( 254 )     

      The new lightweight block cipher ESF is applied in the Internet of Things to provide security for RFID tags and smart cards. ESF algorithm has a 32round generalized Feistel structure, and its round function employs SPnetwork structure. The block size of ESF is 64bit and the key size is 80bit. We search for the impossible differential feature of ESF using the impossible differential cryptanalysis, and present an impossible differential cryptanalysis on 11round ESF, which is based on the 8round impossible differential. It concludes that the full 32round ESF has enough security immunity against the attack.

      An attributebased encryption scheme with attribute#br# revocation and key delegation based on policy control      
      CHI shuiming, CHEN qin, DANG zhengqin
      2013, 35(9): 94-98. doi:
      Abstract ( 169 )   PDF (1129KB) ( 219 )     

      Based on the analysis of existing attributebased encryption schemes with attribute revocation, a new attributebased encryption scheme with attribute revocation and key delegation based on policy control is proposed. The arbitrator of scheme can control attribute revocation and key delegation flexibly via two policy table, i.e. table of attribute revocation and table of attribute proxy. By querying the attribute revocation table, the arbitrator refuse to decrypt for the user of which attribute has revoked to achieve the purpose of attribute revocation. When the decryption right is delegated, the key of delegation conversion is sent to the arbitrator. By using of attribute proxy table, the arbitrator decides if the user has the right of delegation, so as to achieve the goal of flexible attribute proxy. The scheme uses a linear secret sharing matrix to construct the access policies in order to support flexible access structure and applies the key division technology to the key distribution. Finally, the correctness and security of proposed scheme are proven.

      论文
      Survivability based node weight diversity deploy algorithm         
      LIU Wenjian,HUANG Zunguo
      2013, 35(9): 99-103. doi:
      Abstract ( 90 )   PDF (659KB) ( 227 )     

      It is widely believed that diversity in network systems and software will decrease the virulence of malicious codes and the domino effect of singlerepeat attacks. Proposes a survivability based nodeweight diversity deploy algorithm. Starting from node importance, the algorithm assigns distinct software to all nodes. The algorithm can guarantee the minimal impact of a single node on the network systems as well as achieve high survivability of the core nodes. The algorithm is proved to be effective through simulation experiments.

      On the equivalence between deduction systems
      H and G  in propositional calculus     
      SONG Fangmin,YIN Xiyao
      2013, 35(9): 104-109. doi:
      Abstract ( 105 )   PDF (348KB) ( 207 )     

      Firstly, a deduction system G for the propositional calculus in the style of Gentzen is introduced. Secondly, a deduction system H for the propositional calculus in the style of Hilbert is proposed. Thirdly, it is proved that G is equivalent to H in the sense that G and H have the same set of theorems.

      Aspectoriented network software
      evolution method supported by RGPS             
      TIAN Gang1,2,HE Keqing2,SUN Chengai1,CUI Huanqing1,PENG Zhenlian2,3
      2013, 35(9): 110-116. doi:
      Abstract ( 109 )   PDF (1053KB) ( 202 )     

      Pointcuts position definition is not clear and weaving operation is inconvenient are two problems in existing aspectoriented network software evolution method. To address the two issues, an aspectoriented network software evolution method supported by RoleGoalProcessServices (RGPS) process metamodel is proposed. In this method, process is defined by RGPS process metamodel and pointcuts position is defined by metamodel control structure. Aspect pointcuts position definition is precise and can be woven in the original process seamlessly because of the consistent definition of aspect and process. Process evolution is defined using formal method and aspect weaving ways in evolution are clearly elaborated. Finally, a case is used to illustrate the aspect weaving process in the evolution method.

      Service composition based on multiagent learning mechanism  
      ZHAO Haiyan1,2,CAO Jian3,XU Wenbo3
      2013, 35(9): 117-121. doi:
      Abstract ( 79 )   PDF (647KB) ( 203 )     

      Service agent can make use of its own functions and external software services to provide the outside with highlevel services. In order to offer more synthetic functions, service agents should cooperate with each other. A service composition approach based on multiagent learning mechanism is proposed. Its unique feature is that, through learning, agents can improve the speed of solving similar problems after several runs of learning. Firstly, the framework for service composition based on service agents is introduced. Secondly, the multiagent learning mechanism is explained in detail. Finally, the experiment and its results are also given.

      The continuous solution of SAT problem
      based on geometric programming           
      HE Anping,WU Jinzhao,LIANG Yi,XIONG Lingfang,WU Hao
      2013, 35(9): 122-126. doi:
      Abstract ( 96 )   PDF (401KB) ( 190 )     

      The Boolean satisfiability problem (SAT) is fundamental in many fields of computer science; so its fast solution is of great significance. The Solar algorithm is one of the famous quick SAT solutions. We combine the basic Solar algorithm with geometric programming method to propose a novel continuous solution of SAT problem. Experiments demonstrate that the proposal has potential applicable value for solving SAT problems.