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
      A Distributed Trust Proving Algorithm for Open Networking Systems
      WANG Xiaofeng,MA Yanpeng,SU Jinshu
      2011, 33(6): 1-5. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 365 )   PDF (701KB) ( 324 )     

      The traditional trust negotiation methods bring a heavy burden to the trust server and have the blind credential fetching problem. This paper presents a distributed proving and negotiation (DPN) algorithm based on the RTP policy language. DPN can intelligently do a remote trust proving or a local trust negotiation, which can improve the efficiency for trust constructing. Moreover, through recording all the related rules used in the algorithm, DPN can support the verification of the trust proof. Both the correctness and completeness of the algorithm are analyzed, and the experimental results demonstrate the performance improvement of DPN.

      An Efficient Packet Forwarding Strategy Based on the Improved RTS/CTS in MANET
      LU Jia,DOU Wenhua,SHI Shuai
      2011, 33(6): 6-11. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 297 )   PDF (446KB) ( 308 )     

      A mobile ad hoc network(MANET) consists of wireless hosts that may move often. The movement of hosts results in a change in the routes, requiring some mechanism for determining the new routes. Several routing protocols have already been proposed for the MANET networks. Those protocols, based on IP and cost much, affect its stability, extendibility and bandwidth.This paper suggests an approach to modifying the  IEEE 802.11 protocol to enable the hosts to forward packets with the help of the improved RTS/CTS information instead of the routing protocols. It can keep from the effect caused by routing protocols.We have extended the NS2 network simulator to accurately model the forwarding strategy based on the improved RTS/CTS of the IEEE 802.11. The simulation results show that our strategy can effectively reduce the route maintenance overhead and reduce the packet transfer latency.

      A Content Monitoring System Based on Watermarking and Cryptography
      WU Guo1,MENG Qiang2,FANG Liguo1,YI Qingsong3
      2011, 33(6): 12-15. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 412 )   PDF (646KB) ( 296 )     

      Based on an indepth study of watermarking and cryptography, a new monitoring and management program is proposed for monitoring the outdoor LED advertisement content. In this program, the watermark of copyright is bind for the advertisement content before frequencydomain encryption. Under the premise of ensuring the authoritativeness, fairness and practicality, the overall system security is improved and the legitimate rights of the advertising company is safeguarded. A content monitoring and management system of the outdoor LED advertisement is constructed based on the program. Through the establishment of a data center and a monitoring center, the system enhances the management of the outdoor advertisement prior to its broadcast, and the management of the actual content of the advertisement. The outdoor LED can only play the audited content, and perform video recording comparisons to prevent vulnerabilities.

      A Construction Technology of TopologicallyAware Hierarchical Constant Degree P2P
      WANG Xiaohai1,PENG Yuxing1,LI Dongsheng1,ZHANG Honglei2
      2011, 33(6): 16-20. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 327 )   PDF (781KB) ( 268 )     

      The constant degree P2P system has become the P2P domain’s promising hotspot, however, its topologicallyaware problem cannot be resolved by replanting the existing technologies simply. A framework named COFissionE for building topologicallyaware constant degree P2P systems is proposed: the peers are firstly clustered to form the lower level overlay, and at the higher level, a “coincide lower bound” rule is used to construct intercluster links which guarantee efficient intercluster communications and limit the number of intercluster neighbors. The resource publication, query and message routing methods in COFissionE are also provided. The experimental results show that COFissioinE fullfils the  topologicallyaware property with limited overhead and reduces the query cost efficiently. This improvement can be replanted to other constant degree P2P systems with other optimization technologies.

      Design and Implementation of Trusted Paths
      CHEN Songzheng,WEI Lifeng
      2011, 33(6): 21-25. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 306 )   PDF (447KB) ( 321 )     

      The trusted path provides a way for users to authenticate computer systems so that they are assured the systems are not tampered and malicious code such as Trojan Horses couldn't steal their passwords or intercept their sessions. The paper first puts forward a complete design of trusted paths, which aims at Unixlike operating systems and consists of two parts: trusted login and trusted session, and both parts should handle the situations of console interface and graphical interface respectively. And also in accordance with the trusted path, an operating system is divided into four states and a secure attention key will lead to state transitions. With the relation of these states, the design can be more easily mapped into real operating systems. And then the paper gives an implementation through a secure attention key which invokes a trusted path between the user and the system in the FreeBSD operating system. With the trusted paths, FreeBSD can provide a much more secure operating environment for its users.

      Construction of Binary Sequences with FourLevel Autocorrelation
      WANG Yang1,QU Longjiang1,2
      2011, 33(6): 26-30. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 355 )   PDF (366KB) ( 250 )     

      The construction of a family of binary sequences with four level autocorrelation is described. Given any prime p and any positive integers m,n(mn), we construct a family of binary sequences with period pn-1 and fourlevel autocorrelation based on a binary sequence with period pm-1 and threelevel autocorrelation when m|n.The distribution of the autocorrelation values of these sequences is derived. Some properties such as constantonthecoset property and linear span are discussed.

      A New Theory of  Feature Extraction for Binary Images
      CHEN Xuesong,XU Xuejun
      2011, 33(6): 31-36. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 357 )   PDF (1082KB) ( 392 )     

      Extracting the  features of images and targets is the key program in the field of image processing and target recognition. The image potential energy theory is a new method to extract the target features and recognize the target in image processing. This article shows the definition, collection, analysis and application of the binary image potential energy theory. The method shows well the target features, gives us a good result in speed, efficiency, accuracy experimentally. The binary image potential energy theory can be used in target extraction, recognition, tracking and target restoration.

      A Refined Algorithm for the Dominating Sets on Planar Graphs Based on Search Trees
      LAI Xinke,WU Xiaotian
      2011, 33(6): 37-40. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 342 )   PDF (390KB) ( 314 )     

      The dominating set in graphs is considered to be among the most important problems in combinatorial optimization, which is NPcomplete. From the viewpoint of FixedParameter Tractable, the problem is known to be W[2]complete for general graphs, which means that it is impossible to design an algorithms solve the problem with running time f(k)no(1).We investigate the version where we are given a planar graph G=(V ,E), a parameter k, and ask for a set of vertices of size at most k that dominate all other vertices. It is known to be fixedparameter tractable when restricted to a planar graph. We also give a search tree algorithm based on the two sets of reduction rules and the experiments to show the efficiency between different sets of reduction rules.

      A RealTime Visualization Method for LargeScale Terrain Based on GeoMipMaps
      LI Xuemin,LIU Fuyan,YI Song
      2011, 33(6): 41-45. doi:
      Abstract ( 454 )   PDF (923KB) ( 398 )     

      In order to resolve the problem that a massive terrain dataset can not be wholly loaded into the memory in a lump, an efficient method for realtime visualizasion of the largescale terrain environments is proposed. This method improves the GeoMipMaps algorithm and uses a terrain data block technique and a multithreaded technique to implement the dynamic scheduling of.To gain a realtime rendering result, the LOD technology and the frustum culling technology are used to reduce the number of triangles; and the VBO technology is utilized to avoid frequently transmitting large amounts of triangle data from memory to the video memory at the rendering time by storing data which is often used and less frequently changed in the video memory. The experimental results demonstrate that the proposed method can improve the efficiency and visual effect of terrain walkthrough, and render largescale terrain in a realtime way.

      A MultiModality Fusion News Video Story Segmentation Algorithm
      WANG Guoying,KOU Hongzhao,LI Tao
      2011, 33(6): 46-50. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 392 )   PDF (654KB) ( 372 )     

      News story segmentation is an important underlying technology for news video retrieval. This paper presents a method for news video story segmentation, which fuses multimodal features including moderator template matching and topic caption frame detection. News video is segmented firstly based on moderator template matching. And then this news video is segmented based on topic caption frame detection.Finally,the method fuses two segmentation results and removes the repeated segmentation.It is shown by experiments that the algorithm can get better results.

      A Visually Clarifying Display Algorithm for Undirected Relation Graphs
      FANG Wenqi,HU Mingxiao
      2011, 33(6): 51-56. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 388 )   PDF (549KB) ( 397 )     

      A visually clarifying display algorithm for undirected relation graphs is proposed in this paper. Given an ordinary undirected relation graph, after it is processed by the novel algorithm determining the new positions of the vertices, it becomes more clear in aesthetic criteria. First, for a given graph, any isolated vertices are deleted, all the connected branches are decomposed. For each connected branch, it is decomposed into several cliques linked as a tree via recognizing the cut edges (bridges). Then the cliques are decomposed into several subcliques via recognizing the cut vertices. Finally, the vertices in the subclique are uniformly located in a circle. The algorithm is featured with convenient implementation, simple model, fast processing, clear output results and easy parallelization.

      Two Kinds of Curves with Shape Parameters
      YAN Lanlan1,2,LIANG Jiongfeng3
      2011, 33(6): 57-62. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 456 )   PDF (369KB) ( 301 )     

      Two kinds of trigonometric spline bases are constructed in this paper. Based on these bases, two kinds of trigonometric spline curves are defined. As each piece of these trigonometric spline curves are generated by three consecutive control points, these curves retain many properties of the quadratic Bspline curve, but they have a higher order of continuity than the quadratic Bspline curve. For equidistant knots, these curves are C3continuous, and they are C5 continuous under special conditions. The shape parameters of the curves have an explicit geometric meaning. The curves approach the control polygon as the parameter increases. Besides, these curves are closer to the control polygon than the quadratic Bspline curve when the shape parameters are under special conditions. By using the tensor product method, the two kinds of curves can be extended to surfaces. The surfaces have a higher order of continuity than the biquadratic Bspline surfaces.

      A Novel DCTBased Video Text Detection Algorithm
      LIU Lingxia1,NIU Honghui1,CUI Zhoujuan2
      2011, 33(6): 63-66. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 387 )   PDF (692KB) ( 334 )     

      To help users navigate the libraries of video, algorithms that automatically index video based on the content are needed. In this paper, we present a DCT based approach to detect texts and captions from the videos. The use of these features is in a flexible manner thus can be adapted to different applications. Language independence is an important advantage of the proposed method. Experiments are conducted on a large volume of real video shots. Solutions are proposed for each of these problems and compared with the existing work found in the literature.

      An Approach to Contour Extraction Based on Connected Regions
      WANG Wenhao,ZHOU Hong,YAN Yunyang
      2011, 33(6): 67-71. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 378 )   PDF (1083KB) ( 553 )     

      Contour extraction is one of the fundamental steps in computer vision. The correctness and reliability of its results affect directly the comprehension of machine vision systems for the objective world. There are many contour algorithms at present, and each has its own characteristics and shortcomings. Therefore, this paper gives an algorithm based on the connected regions and the threshold to not only extract the contour of the object but also erase the noise at the same step in image processing. Furthermore, the object can be located according to the contour. As shown in the experiments, the noises are erased clearly when the area of each noise is less than that of the object, and especially a single pixel width contour of the object can be presented which is also noninterval and noncross, and not distorted from the object.

      AgentOriented Software Design Patterns
      MAO Xinjun,CHANG Zhiming
      2011, 33(6): 72-78. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 352 )   PDF (479KB) ( 288 )     

      Design pattern gives general solutions to the repeatedlyoccuring problems on certain contexts. It has been widely used in objectoriented software engineering and proved to be helpful to improve the quality and efficiency of software development. We believe, the same design pattern will have various design details when adopting different implementation techniques, and different software development paradigms have their design patterns. As a novel paradigm, agentoriented software engineering has made great progress. Nowadays, many focuses have been put on how to improve its practices and to extend its applications. In this paper, pattern approach is integrated with agentoriented software engineering. Based on the characteristics of the agent technology, a number of agentoriented design patterns have been presented from multiple viewpoints such as structure, collaboration and agent architecture. A description framework for agentoriented design pattern is presented. An analysis of the typical agentoriented design pattern and its application case is conducted.

      Research of the Hot Paths in Software Automation Testing
      MU Yongmin,JIANG Yu,ZHANG Zhihua
      2011, 33(6): 79-83. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 322 )   PDF (727KB) ( 326 )     

      In pathoriented software testing, large systems generate vast amounts of static paths so that it is difficult to test all the paths for testers. This paper proposes the idea of hot path, by using this idea, it can quickly find the paths which easily generate flaws in many static paths. The tree structure of Hot Function can quickly locate Hot Spots and show them. It can bring more convenience to software testers, to improve the testing efficiency and to reduce the cost of testing.

      A Memory Model Based on Abstract Symbol Tables
      DAI Ziying,MAO Xiaoguang,MA Xiaodong,WANG Rui
      2011, 33(6): 84-90. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 419 )   PDF (436KB) ( 367 )     

      Symbolic execution plays an important role in the area of software testing and program verification. However, there are several difficulties facing symbolic execution, one of which is how to abstract various data types and syntax in the source codes. This paper addresses this problem by proposing a new concept of abstract symbol table and a method to model memory using abstract symbol tables. The abstract symbol table records names, types, abstract addresses and symbolic values of addressable objects, which is a simple and accurate memory abstracting mechanism. The memory model is prerequisite for any techniques involving symbolic execution, but this paper systematically presents a memory model for symbolic execution in detail. The abstract symbol tablebased memory model can handle various data types and syntax uniformly including function and class, handle the aliasing problem directly, and possess good scalability because of several performanceimproving techniques.

      A  ComponentBased Software Architecture of Flight Simulation Systems
      ZHANG Zhichun,BI Jianxin,XU Kun,LI Xiaoqi
      2011, 33(6): 91-96. doi:
      Abstract ( 471 )   PDF (575KB) ( 354 )     

      The simulation systems are large, have stringent realtime fidelity requirements, have long lifetimes and usually need to be developed by distributed teams. Those features request that the simulation software must have three essential qualities: good performance, integrability and modifiability.This paper presents a generalpurpose software architecture to achieve that goals. The software architecture consists of the executive subpart and the application subpart that need to be filled in with the actual simulator functionality for making a real flight simulator.The pattern has three time management schemes: a periodic time management, an eventbased time management and a mixedtime system . The architecture is so simple that any entire flight simulators can be completely described by only six module types . Some simulators have been made already by fleshing out this skeleton with the special software requirements such as performance, integrability and modifiability.

      Research on AspectOriented Architecture Modeling
      WANG Rui,MAO Xiaoguang,DAI Ziying,WANG Yanni
      2011, 33(6): 97-101. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 370 )   PDF (478KB) ( 295 )     

      Aspectoriented architecture modeling is an important part of aspectoriented software development and a hot topic in the field of aspectoriented research. The traditional software architecture design approaches do not provide an independent mechanism for crosscutting concerns. As a result, it is necessary to provide a new mechanism for crosscutting concerns in architecture modeling. This paper proposes a concept framework for aspectoriented architecture; then the architectural design phase and the core aspectoriented concepts in the concept framework are considered jointly. As a result, we propose an aspectoriented architecture modeling approach which introduces the aspectoriented concepts to the software architectural design phase.

      Research on High Resolutional Numerical Computation
      ZHANG Xiaoxia,HAO Yizheng,SHAO Jingyun,YUAN Guoxing
      2011, 33(6): 102-107. doi: 10.3969/j.issn.1007130X.2011.
      Abstract ( 370 )   PDF (457KB) ( 280 )     

      High resolution is an important question in high confidence computing. As compared to the conventional numerical computing, high resolutional computing requires high performance supercomputers and high quality programs.The development of parallel computers makes it possible for large scale scientific computation,espically for the boost of resolution in numerical computations.At the same time,a boost in resolution brings new and high requests in the computer’s capability/computational method/physical modeling and parameters. Based on the research of a twodimentional hydrodynamical computational problem, we investigate the key technologies(such as initial priming area/time step/artificial glutinosity/computer’s simulative error/boost of computing quantity) in the boost of resolution, solve these questions and also improve the precision of computation.