Computer Science

Thoughts on Software Testing

NIE Chang-hai

Computer Science. 2011, 38 (2): 1-3.

Abstract

PDF(377KB) ( 999 )

RelatedCitation | Metrics

In recent years, software testing is getting more and more attention from Industry, Education and Academic.Combining the author's experience in teaching and research in this area, this paper presented three level thoughts on the course, professional and science of software testing respectively, discussing on what constitutes the knowledge body as a software testing course,what is vocational skills and literacy of software testing as a professional,and what is the major scientific problems as a science.

YHSSAS: Large-scale Network Oriented Security Situational Awareness System

JIA Yan,WANG Xiao-wei,HAN Wei-hong,LI Ai-ping,CHENG Wen-cong

Computer Science. 2011, 38 (2): 4-8.

Abstract

PDF(597KB) ( 1599 )

RelatedCitation | Metrics

NSSAS(Network Security Situation Awareness System) can perceive the overall network security situation.It plays an important role in improving the emergency response capacity and active defense capability of the networks.Existing NSSAS is limited in processing capacity,which is difficult to apply to the national backbone network and other larg-scale networks. This paper introduced a NSSAS designed for larg}scale networks named YHSSAS. The situational awareness model was proposed first, and then several key technologies, including data integration, correlation analysis,indicators quantification and event predication,were detailed.

Research on Text Information Processing:Review

YUAN Ding-rong,ZHONG Ning,ZHANG Shi-chao

Computer Science. 2011, 38 (2): 9-13.

Abstract

PDF(524KB) ( 1172 )

RelatedCitation | Metrics

The object of text information processing is to analyze and transact information about text from superficial to interior, here to there. It includes extracting information from the text file, getting some meaning hidden in the text by analyzing and mining it, We discussed the work about text information processing in detail,and summarized the history of text information processing. We summed up text information transaction into word segmentation, information extracting from text, text classification, text information retrieval, automatically summary from text etc, and summarized the correlation research work and pointed out the problem in the fields. At last we discussed the key technology and challenge in the text information processing and pointed out that the object is text semantics analysis, induction, deduction and conversion between text and language.

Advances of Support Vector Machines(SVM)

GU Ya-xiang,DING Shi-fei

Computer Science. 2011, 38 (2): 14-17.

Abstract

PDF(384KB) ( 2110 )

RelatedCitation | Metrics

Support vector machines(SVM) arc widespread attended for its excellent ability to learn, that arc based on statistical learning theory. But in dealing with large-scale quadratic programming ( QP) problem, traditional SVM will take too long time of training time, and has low efficiency and so on. This paper made a summarize of the new progress in the SVM training of algorithm, and made analysis and comparison on main algorithm, pointed out the advantages and disadvantages of them,focused on new progress in the current study-Fuzzy Support Vector Machine and Granular Support Vector Machine. Then the two mainly applications-classification and regression of SVM were discussed. Finally, the article gave the future research directions on SVM prediction.

Overview of Technologies for Parallel Component

PENG Yun-feng,YAO Lin,ZHAO Chong-chong,HU Chang-jun

Computer Science. 2011, 38 (2): 18-27.

Abstract

PDF(993KB) ( 990 )

RelatedCitation | Metrics

Parallel component technology is an important way to raise the productivity of parallel computing software and it is an important research area of parallel software engineering. Parallel component model, parallel component architecture and parallel component framework arc three aspects of parallel component technology. Definitions and realizalions of components,interfaces and component frameworks are the most important differences between different archilectures. Much research has focused on improving the performance of parallel applications while realizing the reusing of components. Single program multiple data and multiple programs multiple data programming models, language interoperability, parallel remote method invocation, solutions to MxN problem, component model interoperability emerged in that research. The trends on parallel component technology include research on language interoperability,predictive performance, component adaptation, component model interoperability and parallel component technologies on multi-core.

Trust-Attribute-based Access Control in P2P Environments

FENG Xiao-sheng,WANG Zhen-wen,LI Xiang-yun

Computer Science. 2011, 38 (2): 28-31.

Abstract

PDF(482KB) ( 673 )

RelatedCitation | Metrics

Traditional access control models which are based on identity are not adaptive in P2P environments, which characterize non-centralization, autonomy and dynamic characteristic. We analyzed the access control issue in P2P environments. The existing trust based role access control lacks measures to distinguish users whose results from a trust model are same. We proposed trust attribut}based access control to deal with this problem. Trust attribute-based access control describes users and resources using user attributes and resource attributes. The model builds user role assignment using user attributes, the result from a trust model, environment attributes and authorization policy and builds role permission assignment using resource attributes and authorization policy.

Research on Network Security Policy Refinement Consistency of Detection and Conflict Resolution Mechanisms

NI Jun,CHEN Xiao-su,LIU Hui-yu,LI Jing

Computer Science. 2011, 38 (2): 32-37.

Abstract

PDF(561KB) ( 1324 )

RelatedCitation | Metrics

Through policy-based network security management research, this paper analyzed the existing network security policy conflict detection and resolution method shortcomings. Based on policy refinement of ideas and security policy conflicts classification technology, policy-based network management security-level model was established, with extended XACMI_ language description. According to the relationship between policy behavior, using knowledge reasoning,dynamic layered security corresponding level of policy refinement consistency automatic detection and timely conflict resolution were made, letting it has a good reusability and scalability, and is conducive to the improvement of management efficiency. Policy-based access control refinement application implementation was verified. Finally, some of the fulure research directions were discussed.

Delay Analysis for GBN-ARQ Protocol over Infrared Communication Systems

LI Suo-ping,TENG Guo-bao,ZHOU Yong-qiang,MA Cheng-ye

Computer Science. 2011, 38 (2): 38-41.

Abstract

PDF(357KB) ( 726 )

RelatedCitation | Metrics

Because of the specific property of half-duplex over the IrDA links, Go-Back-N Automatic Repeat request protocol(GBN-ARQ) is different from GBN-ARQ over duplex links. In this paper, the GBN-ARQ protocol over half-duplex links was proposed, and the factors which affect the transmission system were expressed with proper parameters according the property of half-duplex links. I}hen, the general expression of average packet delay was obtained through analyzing the component of average packet delay by using queuing theory, which contains average packet waiting delay and packet service delay. At last, the influence of packet error rate, windows length and data packet length on system delay was analyzed by numerical simulation.

Research on Information Brittleness Structure in Network Communication System

XUE Ping,XU Jun-feng,LIU Yang

Computer Science. 2011, 38 (2): 42-45.

Abstract

PDF(436KB) ( 622 )

RelatedCitation | Metrics

From system brittleness research point of view, this paper defined all the factors and incidents in a system,which can influence the system and arouse the system brittleness, as brittleness environment These factors and incidents can cause sudden collapse of the system. Based on the brittleness of the system environment, the paper established the network communication system brittleness structure model. Combined with brittleness entropy, process of information brittleness risk in a network information system was analyzed. The result of the analysis shows the practicable of the method. The effectiveness and stability of the system information brittleness structure analysis method were proved in the paper. In the end, through building analysis model of main brittleness factor, the paper applied the analysis method to a specific network information system, and obtained satisfactory results.

Identity-based Broadcast Encryption Based on Ad hoc Networks

SUN Jin,HU Yu-pu,ZHANG Le-you

Computer Science. 2011, 38 (2): 46-49.

Abstract

PDF(350KB) ( 768 )

RelatedCitation | Metrics

To give concurrent consideration in the security and efficiency, a novel identity-based broadcast encryption was developed for ad hoc networks. Through security analysis, the correctness and effectiveness of the above methods were verified. Based on the standard model, the bilinear maps were introduced to achieve the collusion-secure for arbitrarily large of users during the broadcast communication. In the process of extract, new users could join dynamically without modification of user decryption keys nor ciphertext size by using ID,and its size not much than O(1). Simultancously, all efficiency measures were completed in O(r) time but not depended on the number of user m, thus the efficiency of transmission and storage was improved. Proof of security shows that the proposed scheme is IN-IN-CPA secure against static adversaries in the standard model under the (t,n)-GDDHE intractability assumption.

Fair and Anonymous Mobile Payment System Utilizing Smart Card

CHENG Jie-ren,CAI Zhi-ping,YIN Jian-ping,ZHANG Ling

Computer Science. 2011, 38 (2): 50-54.

Abstract

PDF(511KB) ( 670 )

RelatedCitation | Metrics

A fair and anonymous mobile payment system model utilizing smart card was proposed. An anonymous offline digital-cash payment protocol was presented. It uses smart card as a distributed anonymity provider agent based on Schnorr signature scheme. It provides dynamic accountable anonymity and changing for digital-cash. It was proved that secure and accountable anonymity of digital-cash was guaranteed. The protocol was proved efficiency in mobile environmenu. A protocol for fair exchange timcsensitive goods was proposed. Our approach utilizes smart card on the Gustomer's side partly taking over the duties of the intermediary party. Besides supporting disconnected operations, our approach also allows the proper handling of time-sensitive items,a feature that previous protocols lack.

Process Expression of Stable Bounded Petri Nets

WANG Ming-xin,LIU Guan-jun,YAN Chun-gang

Computer Science. 2011, 38 (2): 55-58.

Abstract

PDF(379KB) ( 651 )

RelatedCitation | Metrics

The paper[1]proved that the process expression of a bounded Petri net is a regular expression with the basic subprocess set of the net system as alphabet. However, methods for the solution of the basic subprocess set have not been proposed so far. This paper defined a class of bounded Petri nets that are named by stable bounded Petri nets(SB-PNs). A method was presented to compute the set of basic subprocesscs, and then the process expression of an SBPN could be produced by a finite state machine with the basic subprocess set of the net system as alphabet. Next, an algorithm was proposed to construct the process expression of a Petri net synchronized by a set of S-nets.

Cross-layer Routing Design for Wireless Sensor Networks Based on Multiple Ant Colony Algorithm

LI Li-fen,ZHANG Jun-yan,ZHU Yong-li,YU Yong-hua

Computer Science. 2011, 38 (2): 59-62.

Abstract

PDF(464KB) ( 596 )

RelatedCitation | Metrics

Since the event driven wireless sensor networks required strong real-time and high reliability for data transmission, the access efficiency and queue length which were obtained from the media access control layer were regarded as routing metric parameters. However, there is increasing concern that the AntNet algorithm is being disadvantaged because of large transmission delay and slow polymerization in the heavy-loading wireless sensor networks, so the loadsensing and efficient accessing routing protocol was proposed based on multiple ant colony algorithm. In order to seek the optimal path for data transmission in wireless sensor networks, the delay, load, access efficiency were considered as routing metric parameters. The results show that the cross layer routing protocol can guarantee the real time for data transmission and achieve the congestion control and load balancing for wireless sensor networks.

Optimization for Data De-duplication Algorithm Based on Storage Environment Aware

ZHOU Jing-li,NIE Xue-jun,QIN Lei-hua,LIU Ke,ZHU Jian-feng,WANG Yu

Computer Science. 2011, 38 (2): 63-67.

Abstract

PDF(456KB) ( 693 )

RelatedCitation | Metrics

Storage applications such as backup and archive are creating more and more duplication data, which has caused increasing waste in storage space and energy consumption, and how to delete duplication data has become a hot subject in research. CI}C(Content-Defined Chunking) is a prevail algorithm for data dcduplication and can be applicable in various environment,however it does not take into account some characteristics which are specific to individual environment and can influence its result. We studied the CDC' s application in storage system and put up with two constraints for determining parameters for CDC; (1) Determining parameters such as average block size based on the block organization of storage devices; (2)Determining block boundary based on candidate boundary distribution. The result indicates that, comparing with the separate CI}C without restraint conditions, these two constraints can achieve 16. 3% higher compression ratio with 4 data sets.

Research on the Implementation of Intelligent Handoff Technique of VoIP Based on RTP/RTCP

LI Qin-de,ZHOU Wen-an,MA Fei,SONG Jun-de

Computer Science. 2011, 38 (2): 68-71.

Abstract

PDF(380KB) ( 698 )

RelatedCitation | Metrics

As to satisfy users' different demand with the telecommunication, access networks' heterogeneity has been the inevitable trend of next generation networks. To solve the bad continuity of the telecommunication caused by not only the unstable performance of the heterogeneous network and terminals but also the real-time change of users' demand,we proposed a user-oriented and RI’P/RhCP based implementation of intelligent handoff technique of VoIP. By means of the extension and application of RTP/RTC;P protocol, we can detect the real-time end-to-end network QoS performance in VoIP service, and collect the information of users and the performance of terminals and networks, so the system could utilize the performance thresholds of the service, the network environment, the performance of terminals and users' preferences to intelligently handoff users to the best network and terminal,meeting the demand of the service contmmty and access diversity.

Application of Improved Shuffled Frog Leaping Algorithm in Optimum of Sensor Location

LIU Xiao-qin,HUANG Kao-li,AN You-lin,LU Xiao-ming

Computer Science. 2011, 38 (2): 72-75.

Abstract

PDF(399KB) ( 604 )

RelatedCitation | Metrics

Optimum of sensor location is an important research field in testability design, and it is a new attempt to use shuffled frog leaping algorithm for optimum of sensor location. Considering the optimal problem of sensor location is set in a space featuring discrete, a discrete shuffled frog leaping algorithm was proposed, and the change in position was redefined discretely. To avoid converging too fast, the algorithm was improved. Chaos optimization algorithm was used to optimize the best solution in the form of probability. An example and simulation results were provided to verify the effcctivcncss and practicability of this approach.

Improved Study Based on DV-Hop Localization Algorithm

WANG Xin-sheng,ZHAO Yan-jing,LI Hai-tao

Computer Science. 2011, 38 (2): 76-78.

Abstract

PDF(351KB) ( 832 )

RelatedCitation | Metrics

DV-Hop is a typical rangcfrec localization algorithm for wireless sensor network. Estimably calculating the average hop-distance is one of the most important factors which decide the final localization precision. The average hop-distance estimated only by the nearest anchor node was used in classical DV-Hop localization algorithm, so it brings larger localization error. In order to improve the localization accuracy of the nodes, an improved algorithm based on the original algorithm was proposed. The average hop-distance which is estimated by multiple anchor nodes was considered to use and weighted average hop-distance was introduced instead of the average hop-distance which is used in the classical algorithm. The simulation results indicate that the novel algorithm reduces the localization error caused by the calculation of hop-distance effectively, and improves localization accuracy at the same network environment.

A Kind of Authentication Routing Protocol Based on Double Satellite Network in Space Information Network

HAO Xuan-wen,MA Jian-feng,REN Fang,LIU Xiao-yue,ZHONG Yan-tao

Computer Science. 2011, 38 (2): 79-81.

Abstract

PDF(275KB) ( 754 )

RelatedCitation | Metrics

The concept was introduced about of Space Information Network. I}he present condition for research on satcllife network routing and the routing needs for satellite network topology were analyzed with space information network environment as background. Based on the routing safety needs for satellite network, the satellite network routing protocol was studied and an authentication routing protocol, based on double satellite network, was proposed. Furthermore,the safety of the authentication routing protocol was proved. The research finds that the safety of satellite network routing information is significantly advanced.

Adaptively Extracting Application Signatures from Session

WANG Bian-qin,YU Shun-zheng

Computer Science. 2011, 38 (2): 82-85.

Abstract

PDF(467KB) ( 669 )

RelatedCitation | Metrics

Accurate identification of application layer traffic is important for many Internet applications, such as provision of differentiated services, quality of service(QoS) guarantee, intrusion detection, traffic monitoring, accounting, and so on. However, there is no effective method which can automatically extract application signatures. In this paper, a novel method based on frequent item mining was presented,which can automatically extract frequent set from sessions of any application protocol, reduce redundancy of the frequent set based on adaptive filtering rules, and obtain the application signatures. Identification rate and precision rate were applied to verity the extracted signatures. Experiment results show that this method is effective and the extracted signatures are subtle. Therefore it can be used to accurately identify the corresponding application.

Multiple Node-disjoint Paths Distance Vector Routing for Ad hoc Networks

GUO Xian,FENG Tao,YUAN Zhan-ting

Computer Science. 2011, 38 (2): 86-90.

Abstract

PDF(464KB) ( 609 )

RelatedCitation | Metrics

The AMR algorithm used in computing the maximal set of multiple paths is an aggregated multi-path routing scheme for Ad hoc Networks. The MDVR protocol proposed in this paper is a derivative of AODV integrated with the AMR algorithm for the discovery and maintenance of multiple paths. The length of route packets can be shortened and the burden of source for route discovery can be lightened, because paths established by MDVR are saved in routing lable. In addition, the attacks for MDVR by modifying paths in route packets can be efficiently avoided.

Design of DSI for LTE MBMS Service

WU Wei-min,JIANG Xiao-wei,ZHU Guang-xi,GAO You-jun

Computer Science. 2011, 38 (2): 91-94.

Abstract

PDF(353KB) ( 671 )

RelatedCitation | Metrics

Dynamic scheduling has important theoretical and practical significance for improving the resource utilization ratio of MI3MS(multimedia broadcast/multicast service) , and the design of dynamic scheduling information(DSI) is animportant part of it. The paper discussed some problems existing in current DSI design and their impact on user's reception of MI3MS traffic,and built an OPNET simulation platform which is strictly based on the 3UPP standard traffic models and I_TE protocol stack to study some of these problems. According to the simulation results, we presented an more practical and bit-saving DSI design proposal using differential sulrframe offset to denote the dynamic scheduling information.

On Compatibility of Asynchronously Communicating Web Services

LIU Fang-fang,YU Jie,LI Cheng-ling

Computer Science. 2011, 38 (2): 95-99.

Abstract

PDF(430KB) ( 578 )

RelatedCitation | Metrics

Compatibility of Web services is an important issue in the research field of Web services. In Web environment, Web services from different organizations communicate asynchronously. In this paper, based on formal description of Web services and services composition, we analyzed the constraints of compatibility of services and developed the method to check compatibility of component services.

Cryptanalysis of Rabbit

ZHANG Zhen-guang,HU Yu-pu,WANG Lu

Computer Science. 2011, 38 (2): 100-102.

Abstract

PDF(228KB) ( 738 )

RelatedCitation | Metrics

By analyzing the weakness in design of the stream cipher Rabbit, we presented key-recovery attack on it, Afto guessing 96bits, 96bits and 5bits in turn, we can obtain the internal variables, the counter variables and the secret keys in sequence. The result shows that precomputation complexity of whole process is O(296)time complexity is O(297)，request memory space is O(295.81).Comparing with existing attack algorithm, it adds precomputation complexity and memory space, but reduces time complexity.

Inframarginal Analysis Based Resource Allocation Method in Distributed Computing

LIU Jun,WANG En-ze,QIAO Jian-zhong,LIN Shu-kuan

Computer Science. 2011, 38 (2): 103-105.

Abstract

PDF(347KB) ( 678 )

RelatedCitation | Metrics

The key of distributed computing is to fully utilize computing resources. However, the efficiency of distributed computing can always be affected by volatility of available nodes and uncertainty of network environment In order to enhance resource allocation efficiency in distributed computing, this paper presented an inframarginal analysis based resource allocation method, which distributes computing task to the node with more comparative advantage carrying on the task. Simulation results proved that it is effective on resource allocation in the distributed computing environment.

Public Traceable Traitor Tracing Scheme Secure against CCA2

WANG Qing-long,ZHANG Shao-bo

Computer Science. 2011, 38 (2): 106-109.

Abstract

PDF(443KB) ( 664 )

RelatedCitation | Metrics

A new k-resilient publi}key traitor tracing scheme was proposed. A traitor in this scheme can be traced by the way of public black-box tracing mean. Meanwhile, our scheme is provably secure against adaptive chosen ciphertext attack assume that DDH problem is difficulty. Our scheme remains CCA2-secure when not more than k traitors arc reyoked. Compared with similar schemes,our scheme does not use one-time message authentication code assumption and reduces the tracing complexity greatly. Further, our scheme satisfies asymmetry.

Branch-bound Algorithm for Undirected Time-depended Chinese Postman Problem

TAN Guo-zhen,SUN Jing-hao,XIAO Hong-ye,LU Kai

Computer Science. 2011, 38 (2): 110-113.

Abstract

PDF(345KB) ( 751 )

RelatedCitation | Metrics

The time-dependent network Chinese postman problems(丁DCPP) have more applications than the classical Chinese postman problem(CPP). Many transportation and communication systems can be represented by network with travel times that arc timcdependent. When the arc traveling time of system model is timcdependent, the problem becomes considerably more difficult Firstly, it was proved that the standard CPP algorithms can not be used to solve the TDCPP. Then according to the properties for the TDCPP, a branch and bound algorithm for undirected timcdepended Chinese postman problem was derived. While, particularly, for the FIFO time varying network, a branch and bound algorithm was proposed which can solve the problem more effective. Finally, experiment was implemented on micro computo and results show that the algorithm can solve the TDCPP successfully.

Dynamic Prediction Method for Web Service QoS Based on Case-based Reasoning

LIU Zhi-zhong,WANG Zhi-jian,ZHOU Xiao-feng,LOU Yuan-shang,SHANG Ling

Computer Science. 2011, 38 (2): 119-121.

Abstract

PDF(387KB) ( 562 )

RelatedCitation | Metrics

With the rapid growth of functionally similar Web services over the Web,Quality of Services(QoS) is becoming a decisive factor for Web service selection. However, current QoS measurement method generally takes the mean value of Web service history QoS values as the service's QoS, and hasn't taken the dynamic nature of service performance into consideration, and cannot measure QoS of Web services accurately, as a result, many selected services cannot satisfy consumer' s QoS constraints. In our framework, a dynamic QoS prediction method based on Cascl3ased Reasoning(CI3R) was provided, this method associates Web service QoS with the environment of service, task type and task size together, when a new service request comes, CI3R is applied to predict QoS of the Web service in completing the new request Experimental results show that, this method can improve accuracics of QoS of Web service effectively.

Research on Software Interfaces of Integrated Avionics System

CUI Xi-ning,HU Lin-ping,YE Hong,BAI Xiao-ying

Computer Science. 2011, 38 (2): 122-126.

Abstract

PDF(471KB) ( 1450 )

RelatedCitation | Metrics

Avionics system rectuires aviation tasks to obtain more comprehensive capabilities. A deep analysis was made on the requirements of avionics system software, the definition of the Framework for Evaluating Avionics System Capabilities was proposed. The comprehensive capabilities evaluating was firstly introduced into the integrated avionics system, and the framework of the capabilities evaluating was established. Based on these, the analysis and comparison of the current interface specifications were presented and the evaluation result of these specifications was illustrated. It is benefit for the improvement of the reusable,transplantable,secure and available capabilities of the avionics system software to combine the merits of the several specifications.

Study on Software Defect Patterns

ZENG Fu-ping,JIN Hui-liang,LU Min-yan

Computer Science. 2011, 38 (2): 127-130.

Abstract

PDF(350KB) ( 1790 )

RelatedCitation | Metrics

Software defects are the root causes that lead software unreliable, thus the reduction of software defects is the key way to improve software reliability. How to improve the software reliability using the defect data accumulated? In this paper, the definition of software defect pattern was addressed on the basis of the concept of software defect and pattern. Through analyzing the defect data accumulated, the classification of defect patterns was proposed. The defect patterns in software requirement analysis, design and coding phase were advanced. How to apply the defect patterns is claborated in the process of software development and testing, which provides idea on how to improve software reliability utilining defect data.

Fault-tolerant Service-oriented Architecture Description Language Supporting Exception Handling

LU Guo-bin,WANG Quan-yu,LI Gui-long

Computer Science. 2011, 38 (2): 131-137.

Abstract

PDF(613KB) ( 555 )

RelatedCitation | Metrics

To solve the problem of the construction of fault tolerant scrviccoricnted software architecture, the paper proposed SOADL-EH, a new service-oriented software architecture description language (ADL) supporting exception handling. With the capability of modeling normal SOA and the provision of exception handling services(EHService)，exception handling connectors(EHConnector) and exception handling configurations(EHConfigure)，etc. software archilecture elements, this language can separate clearly the exception handling logic from the normal business logic of the service-oriented software architecture layer, which covers the shortage in exception handling capability of the existing ADLs.

Research on the Semantic Web Oriented Method for the Evolution of Composite Service

WANG Xiao-xuan,BAO Ai-hua,MIAO Jia-jia,DING Ke,WANG Zhen

Computer Science. 2011, 38 (2): 138-143.

Abstract

PDF(720KB) ( 577 )

RelatedCitation | Metrics

To adapt composite services with changes in dynamic environment agility and effectively,a semantic Web oriented method for the evolution of composite service which named EM4CS was proposed systematically. Taking composite services described by OWI= S as research object, EM4CS divides the evolution process into 3 parts, and covers the evolulion cycle in 6 stages, i. e. evolution requirements capturing, formal representation of evolution requirements, syntactic consistency maintenance, semantic consistency maintenance, confirmation of evolution results, publication and broadcasting of evolution results. The composite service evolution process was analyzed from the perspective of system enginecring, and the evolution requirements were fulfilled in an iterative refined way, in which the syntactic and semantic consistency were also maintained. The composite service evolution support system, which follows EM4CS, named ESS4CS,was also introduced in this paper. ESS4CS can assist knowledge engineer in the whole process of composite service evolution, and the efficiency and precision of the evolution process are also improved.

Simulation of GMPLS-based OBS Network

REN Hai-lan,LUO Zhi-hui,LIU De-ming

Computer Science. 2011, 38 (2): 144-118.

Abstract

PDF(527KB) ( 698 )

RelatedCitation | Metrics

OBS(Optical Burst Switching) is a transitional solution from electrical switch to packet switch, G-LOBS(GMPLS-based Optical Burst Switching Network) is a combination of MPLS(Multi-Protocol I_abcl Switching) technology and OBS technology,which is an effective solution for the evolvement of present optical network to optical internet. The idea of building a uniform G-LOBS network simulator was put forward. The signaling module of RSVP-TE(Resource Reservation Protocol-Traffic Engineering) was designed, and the simulation analysis of RSVP-TE by OPNET was done.

Study of Improved Component-based Software Reliability Model Based on Route

ZHANG Wei,ZHANG Wei-qun

Computer Science. 2011, 38 (2): 148-151.

Abstract

PDF(344KB) ( 548 )

RelatedCitation | Metrics

Component based software reliability analysis typically takes the reliability of component as an invariable property of component itself , while disregards the fact that it is changed when the component context that it interacts with is changed due to a different route. This paper proposed a improved component based software reliability modelbased on route, which introduces the dynamatic transition graph to build the relationship between the route of component based software and component reliabity, and gave a improved method based on route to caculate the componet based software reliability. At last, the result of case demonstrates the feasibility and validity of the approach.

Efficient Web Software Testing Method

QIAN Zhong-sheng,MIAO Huai-kou

Computer Science. 2011, 38 (2): 152-155.

Abstract

PDF(446KB) ( 590 )

RelatedCitation | Metrics

Testing Web software faces great challenges. Starting from constructing the PFD(Page Flow Diagram) for Web software, this work proposed a test path generation approach,which was illustrated by the SWI_S(Simple Web Login System) as an example and presented an effective Web testing model for Web software testing. This method provides a meaningful basis for page flow testing technology.

Different Trade Software Quality Evaluation Model and Test Practice

NIE Nan,DENG Lu-juan,XIA Qi-ming,LI Jian-yong

Computer Science. 2011, 38 (2): 156-159.

Abstract

PDF(379KB) ( 633 )

RelatedCitation | Metrics

Aiming at different trade software quality requirement and application problems, it set up their corresponding quality evaluation and test model, and it described evaluation methods and selection basis related with system quality features. It also provided an application case of engineering measuring software quality evaluation model combined with several engineering measuring software. Test practice process was illustrated,which includes questions and resolutions,such as interface interaction errors and format data type errors. At last, the comprehensive and justice evaluation result was counted up, and software quality could be improved.

M2:An Effective Method for XPath Evaluation

CHEN Rong-xin,LIAO Hu-sheng,CHEN Wei-bin,YE Wen-lai

Computer Science. 2011, 38 (2): 160-165.

Abstract

PDF(553KB) ( 704 )

RelatedCitation | Metrics

The performance of XPath evaluation greatly affects XML applications. This paper presented M2(Matrix Match) method based on node relation matrix query. Relation matrix is constructed from XML region encoding, while XPath evaluation is realized by execution of ctuery primitive sectuence. The navigation feature tends to comply with XPath semantics and support the implementation of reversed axis query and branching query expressed in predicates.The existing relation matrix from the same data can be reused by different queries. I}he loop processing style in matrix construction and evaluation is suit for parallelism Experiments show Mz is an effective XPath evaluation method.

Research on Cluster Aggregation Approaches

YANG Cao-yuan,LIU Da-you,YANG Bo,CHI Shu-zhen,JIN Di

Computer Science. 2011, 38 (2): 166-170.

Abstract

PDF(517KB) ( 1174 )

RelatedCitation | Metrics

Clustering aggregation can offer a partition that could better reflect the inherent structure of the data set by studying and integrating many clustering results of the original data set. Clustering aggregation could detect and deal with the isolated points preferably, which improves the quality of clustering. This paper made an overview of the relevant knowledge of the clustering aggregation, presented the concepts and advantages of clustering aggregation. It presented three approaches to get clustering members according to the ucsd cluster algorithms; analysed their respective advantages, disadvantages and application conditions; presented the existing consensus functions; explained the basic principles and pointed out their limitations. Finally, it discussed the future research directions.

Structure-based Entropy Clustering Algorithm for Heterogeneous Data

LI Zhi-hua,GU Yan,CHEN Meng-tao,WANG Shi-tong,CHEN Xiu-hong

Computer Science. 2011, 38 (2): 171-174.

Abstract

PDF(356KB) ( 778 )

RelatedCitation | Metrics

The dissimilarity measure and clustering approach about the heterogeneous dataset were studied, and a struclure-based entropy clustering SEC algorithm was presented in this paper. Data often do appear in homogeneous groups,the SEC utilizes these structural information to improve the clustering accuracy. Unlike the distribution of numeric data,nominal data are often unbalancedly distributed,whose distribution are often unrelated with their distance measure,due to the above, a new structural information-based entropy computing technology was proposed. By mining the clues in structural information, constructing the weight implying the different distribution information of nominal and numeric attributes, the SEC can automatically identifies the initial locations and number of cluster centriods, and exhibits its robustness to initialization and no iteration in algorithm. Experimental results comparing with other references demonstrate that the proposed method has promising performance.

Research and Application of Ontology-based Intelligent Planning Method

YONG Xi,CAO Cun-gen,BAI Ai-ming

Computer Science. 2011, 38 (2): 175-178.

Abstract

PDF(495KB) ( 689 )

RelatedCitation | Metrics

In order to deal with the shortcomings of classical intelligent planning methods with respect to the utilization of domain knowledge, this paper proposed an ontology-based intelligent planning approach by combining advanced knowledge engineering technology with general-purpose planning algorithm, which could make full use of domain knowledge in ontology-based knowledge base to efficiently resolve planning problems in the domain. "hhe practical method to establish the domain ontology-based knowledge base supporting the planning process was also presented. I}he experiment results indicated that the approach could effectively solve complex application problems and make the planning system actually reach the practical level.

Activity Planning Based on Temporal Constraint NetworK

LI Yong-feng,ZHOU Xing-she,DU Ke-jun,YU Zhi-wen,MAO Rui

Computer Science. 2011, 38 (2): 179-183.

Abstract

PDF(435KB) ( 891 )

RelatedCitation | Metrics

Elders often have difficulties in daily activities because of declining cognitive ability. Assisting elders with information technology has become a new and challenging research area. It's a problem that how to predefining activities into a plan and prompting the elders if they fail to execute it in this area. We gave a more flexible way to present the time constraints of activities based on the traditional way, but as the same time how to resolve the conflicts between activities become a new troublesome problem. I}he concept and theory of "Simple Temporal Constraint Network"(STCN) was then proposed, which will well handle the problem. We presented the time constraints of activities with STCN and checked the STCN's consistency. The "General Checking Algorithm"(GCA) was adopt,but we found that the GCA's complexity will increase by exponentially if the number of activities increases. Then, we gave an improved algorithm to resolve the conflicts between activities and compared the two algorithms. The experiment result shows the ICA has better performance than GCA.

Ontology-based Hierarchical Abstraction Model

WANG Nan,OUYANG Dan-tong,SUN Shan-wu

Computer Science. 2011, 38 (2): 184-186.

Abstract

PDF(361KB) ( 537 )

RelatedCitation | Metrics

The KRA Model offers a generic framework to construct the model of the world W. In this paper, the ontology of concept was introduced in the General KRA model and the Abstraction Object Database was extended to Ontology Class which makes the knowledge sharing and reuse possible in the Perception level, the Language level and the Theory level of the KRA model. The three kinds of ontology abstraction operator(set) were also defined;fundamental operator,entity operator set and connection operator set, working on W and ontology class. Furthermore, the mapping between two different ontology classes was given to realize model abstraction and model reverse-abstraction. The abstraction degrec of ontology class was proposed to represent the degree of ontology class and it is as well pointed out that in such a framework, the models of W with different abstraction degrees can be automatically constructed by using abstraction mapping.

Semantic Web Based Reasoning Model for Controlled Natural Language System

WANG Huan-huan,LI Hu,SHI Yong

Computer Science. 2011, 38 (2): 187-190.

Abstract

PDF(473KB) ( 657 )

RelatedCitation | Metrics

Proposed a semantic Web based reasoning model for controlled natural language system which can be used by experts to describe business logic in controlled natural language. hhe framework of the reasoning model includes two parts,one is language processing part and the other is reasoning part. Firstly, the controlled language sentences were partially made into discourse representation structure through ontology lexicon model based on the WordNet and the controlled natural language interpreter based on ontology lexicon model. Then discourse representation structure was transformed into semantic Web OWL and SWRL through the reasoning part. Finally, semantic Web OWL and SWRL was mapped into Jess facts and rules through template tools, and then reasoning was conducted on controlled natural language according to a predefined axiom of semantic Web. Experiment proves that this mode has greatly enhanced the efficiency of knowledge representation modeling, can basically meet the simple reasoning tasks, and has practical value.

Probabilistic Mixture Model Based Summarization Approach for CWME Discussions

WANG Ai,LI Yao-dong

Computer Science. 2011, 38 (2): 191-194.

Abstract

PDF(471KB) ( 543 )

RelatedCitation | Metrics

To solve the problem of the lack of summarization method of experts' speeches in CWME, a new method was proposed. This method first adopts a probabilistic mixture model to discover latent themes from experts’speeches.Then, it judges whether themes have changed. Once they are changed, this method will generate summary and provide it for experts. The summaries can help experts to inspire each other in CWME. Experimental result shows that the proposed method is feasible and effective.

Property of Divisional Weighted Voronoi Diagram and Area Calculation

MA Li-ling,ZHANG You-hui

Computer Science. 2011, 38 (2): 195-198.

Abstract

PDF(363KB) ( 1540 )

RelatedCitation | Metrics

Divisional weighted Voronoi diagram is based on Voronoi diagram and weighted Voronoi diagram. It can be applied to simulate base station coverage with different sector transmission power in mobile cellular communication.Firstly, this paper defined divisional weighted Voronoi diagram based on weighted Voronoi diagram. It also analyzed property and theorem of divisional weighted Voronoi diagram. Secondly, it analyzed all kinds of area for divisional weighted Voronoi diagram and provided a discrete algorithm to calculate area. Finally, it gave an example of using divisional weighted Voronoi diagrams to simulate the condition of base stations in part of Shijiazhuang city.

Research on Self-localization Methods for Mobile Robots Based on Bayes Filter

ZHAO Zeng-shun,SHEN Ji-bi,WANG Ji-zhen,HOU Zeng-guang,TAN Min

Computer Science. 2011, 38 (2): 199-201.

Abstract

PDF(295KB) ( 1370 )

RelatedCitation | Metrics

This article presented a survey of the most common probabilistic models for self localization algorithm of mobile robot. We proposed a general I3ayesian inference framework which is deduced in detail through a combination of Markov assumption with 13aycsian rule. Under such general framework, we gave a review of the main probabilistic models such as Kalman Filtering Series, Multi-hypothesis Localization, Markov Model Localizations and Monte Carlo localization, etc. , all of which can be captured under this single formalism. This will provide readers a global view of this literature. We emphasized the implementation and drawbacks of Monte Carlo Localization, which is considered as one of the most promising method.

Higher-Order-Cumulant Based Memory Effect Identification of Hammerstein Model

HU Xiao,MA Hong

Computer Science. 2011, 38 (2): 202-205.

Abstract

PDF(315KB) ( 674 )

RelatedCitation | Metrics

This paper focused on memory effect identification of Hammerstein model in Uaussian noise. When input statistics and nonlincarity of Hammerstcin arc unknown, by using higher order cumulant of output signal, two sets linear cquations were proposed to extract coefficients of linear block with memory. Theoretical derivation shows that those two sets of linear equations have unique solutions. I}hey could be used alternately to identify the memory effect of Hammerstein model, and the identification process is not affected by memoryless nonlinear block. Finally, simulations verify that the new developments have higher performance than direct extraction method.

Hierarchical Visualization of Multi-dimensional Forum Data

XU Yan-ru,WANG Chang-bo,LIU Yu-hua,ZHANG Qun-yan

Computer Science. 2011, 38 (2): 206-209.

Abstract

PDF(382KB) ( 573 )

RelatedCitation | Metrics

Visualization of multi-dimensional data is an effective method for helping people understand and analyse multivariate data sets. Forums in Internet have taken a more and more important role in everyone's life. Visualizing forum data will help analyze the pattern of people's participation in forums. This paper first analyzed the feature of forum data such as closely relation with time, mass data size, multi-level data structure, multiple attributes and complex relation within the data. We proposed three methods, such as parallel coordinate, hierarchical map and curve map to visualize the different datasets in network forum. At last, we took data in a real forum for example and showed the results of visua lization.

Hierarchical Structure Based Service Semantic Relationship Discovery

ZHAO An-ping,REN Ke,WANG Xiao-yong,QIU Yu-hui

Computer Science. 2011, 38 (2): 210-213.

Abstract

PDF(353KB) ( 577 )

RelatedCitation | Metrics

Increasingly,application developers seek the ability to search for existing Web services within large Internet based repositories. The goal is to retrieve services that match the user's requirements in a fast and efficient way. It is necessary to explore an approach of service semantic relationship discovery to enhance search engine results with a list of related services for each hit. In this paper, an effective approach for service semantic relationship discovery based on the hierarchical structure was presented. Hierarchical structure is adjusted by moving related services together according to the relevance between scrvices,and the relationships between services can be discovered gradually. Experimental resups demonstrate the feasibility and effectiveness of the approach.

Conceptual Model of Battlefield Environmental(BE) Simulation Based on Ontology

SUN Guo-bing,HUANG Jin-jie,LIU Fei

Computer Science. 2011, 38 (2): 214-217.

Abstract

PDF(341KB) ( 987 )

RelatedCitation | Metrics

In order to describe 13E simulation conceptual models sufficiently and explicitly, a method based on ontology was proposed considering the function of ontology. Based on the study of I3E simulation conceptual models, entities and interaction relationship were given in detail, then I3E simulation conceptual models were presented by OWL, and eight meta-ontology terms were proposed to implement the illustration of 13E simulation ontology.

Weighted Naive Bayes Spam Filtering Method Based on Rough Set

DENG Wei-bin,WANG Guo-yin,HONG Zhi-yong

Computer Science. 2011, 38 (2): 218-221.

Abstract

PDF(369KB) ( 850 )

RelatedCitation | Metrics

Using a classifier based on a specific machinclcarning technique to automatically filter out spam email has drawn many researchers' attention. In a spam filtering process,how to selecting the features of emails and how to design a good filtering algorithm arc two key issues. A new method of features selecting was proposed, which include the head and the other main features of emails. Furthermore, the features' importance degree was measured according to information viewpoint of rough set. With it,a new weighted naW a I3ayes spam filtering was put forward. It can solve the conditional dependence of naW c Bayes efficiently. Simulation results on two email data sets in English and Chinese respectively illustrate the efficiency of this method.

GrM-based Representation of the Concepts and Operations in Rough Set Theory

CHEN Ze-hua,XIE Gang,XIE Jun,XIE Ke-ming

Computer Science. 2011, 38 (2): 222-224.

Abstract

PDF(300KB) ( 593 )

RelatedCitation | Metrics

Different representation for knowledge brings different difficulties in comprehension and programming.Knowledge is regarded as the classification ability according to available knowledge in Rough Set Theory, based on which, a set of algebraic definition for knowledge representation and operation was systematically proposed. However, it is relative difficult to help understand the essence of knowledge in view of algebraic description. Professor Miao in Tongji University built up relations between knowledge and information,he redefined Rough Set algebraic system by information concepts. Furthermore,he also provided equivalent proof for both definitions. In this paper,knowledge and its operation were expressed in form of granular matrix. The equivalent proofs for all three definitions mentioned above were given in this paper.

Comparative Study on Method for Determining Optimal Number of Clusters Based on Affinity Propagation Clustering

ZHOU Shi-bing,XU Zhen-yuan,TANG Xu-qing

Computer Science. 2011, 38 (2): 225-228.

Abstract

PDF(356KB) ( 928 )

RelatedCitation | Metrics

It is crucial to determine optimal number of clusters for the duality of clustering in cluster analysis. Based on Affinity Propagation clustering algorithm, a method for determining optimal number of clusters was proposed to analyze the clustering validity and determine optimal number of clusters by using six clustering validity index These clustering validity indexes were analyzed concretely and the method of using IGP index to determine optimal number of clusters was improved. In connection with eight datasets, the performances of these indexes were compared by simulation experimenu. The results of analysis and experiments show that IGP index is the best to determine optimal number of clusters based on Affinity Propagation clustering.

Research on Semantics Analysis of Strategic Decision Text

ZHOU Sheng,HU Xiao-feng,LUO Pi,LI Zhi-qiang

Computer Science. 2011, 38 (2): 229-233.

Abstract

PDF(441KB) ( 568 )

RelatedCitation | Metrics

Aiming at the low precision rate of video retrieval for virtual television news caused by extracting keywords from strategic decision text,the features and construction modes of strategic decision text were analyzed. The principle of dictionary construction, automatic Chinese words separation, method of function analysis, semantic tree construction and matching approaches between virtual television news video were discussed in detail. Experiments were made on this research. The results show that the precision rate of virtual television news video retrieval can be increased by the employment of semantic tree method.

Fuzzy Dominance-based Rou沙Set in Set-valued Information System

YANG Xi-bei,ZHANG Zai-yue,ZHANG Ming

Computer Science. 2011, 38 (2): 234-237.

Abstract

PDF(304KB) ( 603 )

RelatedCitation | Metrics

The set valued information system was analyzed by dominanccbased rough set approach. By considering the dominance degree in terms of pair of objects, the concept of fuzzy dominance relation was proposed, the fuzzy technique was then employed in the dominanccbased rough set theory. hhe fuzzy dominanccbased rough set model was presented, then corresponding properties were discussed, from which we obtained new theory and practical approach to derive decision rules from the set valued decision system. Some numerical examples were employed to substantiate the conceptual arguments.

Study on Algorithm of Community Detection Based on Information Entropy

WANG Gang,ZHONG Guo-xiang

Computer Science. 2011, 38 (2): 238-240.

Abstract

PDF(273KB) ( 797 )

RelatedCitation | Metrics

There are some faults of present community detection algorithm, which is based on the in degree, out degree and betweenness of nodes,we presented a algorithm based on Entropy to detect community structure. A community ineludes many information and it's Entropy. Members of a community have some common gains or interests, we think that if a member want to join a community, it can't make the entropy of the community exceed a threshold, otherwise it can't be the member of a exist community. Our experiments show the processing and the efficiency of our algorithms.

F- information Camoufladge and Camoufladge-reduction Identification

GENG Hong-qin,ZHANG Guan-yu,SHI Kai-quan

Computer Science. 2011, 38 (2): 241-245.

Abstract

PDF(460KB) ( 633 )

RelatedCitation | Metrics

P-sets(packet sets) are set pair combined with internal P-set XF (internal packet set XF) and outer P-set XF(outer packet set XF) , or(XF, XF) is P-sets. Dynamic P-sets are obtained by improving cantor set, dynamic characteristic of P-sets. I3y using the outer P-set XF，the concept of F-camoufladge carrier,F-camoufladge surplus were given. By using the concept,F- information camoufladge generation and F- information camoufladge structure were given,F- information camoufladge measure was given, F- information camoufladge ring theorem and F- information camoufladgering separation theorem,F- information camoufladge reduction theorem,F- information camoufladge identification were given. I3y using the results, the application of F- information camoufladge in information transfer was given. Information camoufladge is a new branch of studying information system.

P-sets and F-memory Information Characteristic-Application

WANG Yang,ZHANG Guan-yu,SHI Kai-quan

Computer Science. 2011, 38 (2): 246-249.

Abstract

PDF(433KB) ( 522 )

RelatedCitation | Metrics

Packet sets arc set pair combined with internal packet sets and outer packet sets or (XF , XF ) , which possess dynamic characteristic because of attributive deleting and adding of the attributes set a of cantor set X. By using the structure and dynamic characteristic of packet sets, the concept of F-memory information genaration,F-memory information measurement and F-memory circle were given, the F-memory information existing theorem, the F-memory information recovery theorem and the F-memory information characteristic theorem were proposed. I3y using the results, the application of F-memory information was given. Packet sets is a new mathematical model and mathematical method.

P-sets and the Identification-recovery of F-outer Embedding Information

YU Xiu-qing

Computer Science. 2011, 38 (2): 250-253.

Abstract

PDF(392KB) ( 566 )

RelatedCitation | Metrics

By using outer P-sets XF , the concepts of F-outer embedding information (x)F and F-outer embedding infor mation dependence, F-outer embedding coefficient and F-outer embedding degree to information(x)were proposed. Moreover, based on these concepts, existing theorems, dependence theorems and identification theorems about F-outer embedding information were presented. hhe application of F-outer embedding information in information system was discussed.

Simulative Study on HPGL Interpolation Control Algorithms

HUANG Gang

Computer Science. 2011, 38 (2): 254-256.

Abstract

PDF(235KB) ( 570 )

RelatedCitation | Metrics

To smooth the polyline outputted by HPGL format file, the existing disadvantages of data output and corresponding solutions were analyzed. An interpolation processing method based on genetic algorithms was raised, and the mathematical description of the algorithms was given, in addition, the mathematical analysis of the interpolation processing was elaborated. As shown in the simulative experiment, during several interpolation treatments of polyline outputted by HPGI_ format file, as long as the interpolation is controlled in accurate range, the final motion path would become a smooth curve.

Real-time Simulation of Dynamic Cloud Based on Repeating Texture Mapping

WANG Xiang-hai,LI Ting-ting,PANG Xin

Computer Science. 2011, 38 (2): 257-259.

Abstract

PDF(350KB) ( 793 )

RelatedCitation | Metrics

Cloud is a common, widespread natural phenomenon, cloud modeling and rending are a difficult problem in the field of computer graphics. I}his article used Repeat Texture Mapping to make up multiple 2D texture images for modeling 3D dynamic cloud, through the instantaneous translation of the cloud, zooming and other operations to simulate the dynamic changes of cloud, further used alpha color fusion technique to simulate dynamic cloud gather, dissipation and the cloud form changes after the effect of wind, by using the superposition of multilaycr repeat texture mapping to simulate the effect of layer cloud and obtain good simulation of the dynamic effect of cloud. The proposed modeling dynamic cloud method is simple, avoids a large number of tedious calculations, improves the speed of dynamic cloud generation and meets the animated scenarios for dynamic applications such as real-time requirements of cloud. I}he experimental resups validate the correctness of the proposed method of Dynamic Cloud Modeling and rendering generation.

New Efficient Image Fusion Algorithm for Image Mosaic

SONG Bao-sen,FU Yong-qing,SONG Hai-liang

Computer Science. 2011, 38 (2): 260-263.

Abstract

PDF(444KB) ( 1143 )

RelatedCitation | Metrics

Image fusion is one of key technics of image stitch tracer eliminate, multitudinous existing algorithms of impge fusion arc aimed at complete overlap images, but ignore the research for algorithms of partially overlap image fusion for image stitching applications, thus affecting the application and development of image stitching technology. In order to efficiently,high-duality accomplish image stitching, we presented a new method of image stitch tracer eliminate using technology of searching suture silk combined with wavclet multi-resolution analysis. Experiment shows the method not only has a preferable real-time, but also has high-quality result of stitch tracer eliminate.

Fast Color Quantization Algorithm Based on Linked List

YU Qing-jun,YU Bao-chu,TANG Zhen-an

Computer Science. 2011, 38 (2): 264-266.

Abstract

PDF(276KB) ( 639 )

RelatedCitation | Metrics

In order to improve the quality and speed of color quantization algorithm, we took the rapidity of splitting algorithm and the accuracy of clustering algorithm into account during the design process of color palette, and tried to improve the process of pixel matching,and then proposed a new real-time color quantization algorithm. In this algorithm,all colors in an image were connected together via a data structure link according to its appearance frequency. The whole process of color quantization could be viewed as a series of operations about that data structure link. Experimental resups show that the proposed algorithm is able to achieve the expected results.

Sparsity of SAR Image with Application to Feature Enhancement

WANG Guang-xin,LIN You-ming,ZHANG Song-bao

Computer Science. 2011, 38 (2): 267-270.

Abstract

PDF(366KB) ( 908 )

RelatedCitation | Metrics

The sparsity of SAR image data was discussed based on the backscattering properties of targets in the image.Firstly,the paper analyzed the cause of sparsity,and pointed out that the sparsity originates from the backscattering properties of the targets,the sparsity of the strong scattering centers,and the roughness of the imaging area. Secondly,the visual differences of the sparse properties between optical and SAR images were compared based on the sparse representations. Thirdly, the application of these sparsity properties to SAR image feature-enhancement was investigated.Finally, the experiments with simulated data and MSTAR data demonstrate the positive effect of sparse constraints on feature-enhancement.

Robust Watermarking Scheme Based on NMF and SVD

LIU Ru-jing,WANG Ling

Computer Science. 2011, 38 (2): 271-273.

Abstract

PDF(269KB) ( 702 )

RelatedCitation | Metrics

In this paper, a digital watermarking algorithm based on nonncgative matrix factorization(NMF) and singular value decomposition(SVD))was proposed. The host image was decomposed with DWT. NMF and SVD were applied to the low frequency component of the original image. The watermark after Arnold scrambling was embedded into the singular value of the image. The experiment results indicate that the proposed scheme can obtain better visual effect and it is robust against attacks such as added noise, filtering and cropping.

Face Recognition Based on Log-Gabor and Orthogonal IsoProjection

WANG Qing-jun,ZHANG Ru-bo

Computer Science. 2011, 38 (2): 274-276.

Abstract

PDF(363KB) ( 580 )

RelatedCitation | Metrics

In view of the problems of feature extraction in face recognition, a Log-Gabor and orthogonal supervised Iso-Projection based algorithm for face recognition was proposed in this paper. The proposed algorithm first gets the high-order nonlinear statistical information by calculating the Log-Gabor wavelet representation of face images. Then the orthogonal constrained conditions added to the original optimal problem and the iterative formulae for finding a set of orthogonal optimal projection vectors arc deduced. The orthogonal basis can help to preserve the information of nonlinear sulrmanifold space which is related to distance and reconstruct data. The experimental results on ORL and PIE face database illustrate the effectiveness of the proposed method.

Automatic Image Annotation Method and Fast Solution Based on the Mutual K Nearest Neighbor Graph

GUO Yu-tang

Computer Science. 2011, 38 (2): 277-280.

Abstract

PDF(327KB) ( 949 )

RelatedCitation | Metrics

Image semantics has the characters of vague, complex, and abstractive, therefore only low-level features are not enough for describing image semantics, and rectuire a combination of image-related content in order to improve the accuracy of the image annotation. In this paper, an image annotation method based on mutual K nearest neighbor graph (MKNNU) was proposed, which builds the relationship between the low-level features, annotation words and images by a mutual K nearest neighbor graph. Mutual K nearest neighbor graph is to extract semantic information from paired nodes, which overcomes the limitation of unilateral mining of the traditional K nearest neighbor graph, and effectively improves the image annotation performance. Based on the analysis on the structure of mutual K nearest neighbor graph,Combined with Random Walk with Restart(RWR)，a fast algorithm was proposed without apparent reducing the precision of the image annotation. Experimental results on the Corel image datasets show the effectiveness of the proposed approach in terms of quality of the image annotation.

Three-dimensional Geological Modeling Based on Qt and OpenGL

TANG Yun,DENG Fei,HUANG Di-long

Computer Science. 2011, 38 (2): 281-283.

Abstract

PDF(246KB) ( 925 )

RelatedCitation | Metrics

Surface interpolation method was used to interpolate the actual formation discrete control point maped out by the user into smooth surfaces, so as to further obtain a smooth continuous three-dimensional geological model of the geological interface which can be used to express a variety of geological interface. This paper described several commonly used interpolation methods which are suitable for small-scale data and compared the interpolation results, and then introduced hierarchical I3-spline based interpolation method, which can be applied to large-scale scattered data interpolation. Software is based on Qt and OpenGI_ for development, has been very good in engineering geology applications.

Design and Simulation of Lighting and Texture Mapping in GPU

DONG Liang,LIU Hai,HAN Jun-gang

Computer Science. 2011, 38 (2): 284-287.

Abstract

PDF(444KB) ( 763 )

RelatedCitation | Metrics

Graph Process Unit(GPU) usually adopts a pipelined architecture, and implements some general computer graphics API. This paper analysed the working principle and architecture of GPU, and proposed an improved practical pipeline structure. Every module of the pipeline was analyzed. I}his paper focused on the parts of lighting and texture mapping which demand intensive calculations, and presented a design method. The design method has been validated with software simulation and applied in the design of a hardware prototype. In addition, this paper also proposed a programmable architecture for lighting and texture calculations based on the results of simulation study.

Pview: A Novel Implementation of Fundamental Supports for Parallel Programs Performance Monitoring Based on PMU

YANG Jie,XU Heng-yang,AN Hong,LIU Yu,WANG Yao-bin

Computer Science. 2011, 38 (2): 288-292.

Abstract

PDF(485KB) ( 1122 )

RelatedCitation | Metrics

In the past recent years,performance monitoring and profiling have been becoming one of the most important issues in the era of parallel programming. Performance Monitoring Unit(shortly PMU in the rest of this paper) embedded in the modern processor provides a hardware support that makes it possible to monitor a running program online with trivial system disturbance and overhead and further do more post analysis in scrutiny based on the collected data.Pview, meaning performance view, is a novel scheme we designed to fundamentally support performance monitoring for multi-threaded applications in system level, and it is implemented in I_inux 2. 6. 30 platform by extending the kernel.Pview uses an approach of system call to rectuest performance monitoring services, and also co-works with a module called pvicw data collector engine to feed user analyzer tools the collected performance event data. I}his paper presented all of our minds in pvicw design.

RTI. Low Power Technique for SoC Design Using Weighted Datapath

YANG Heng-fu,TIAN Zu-wei,LI Yong-fan

Computer Science. 2011, 38 (2): 293-295.

Abstract

PDF(286KB) ( 638 )

RelatedCitation | Metrics

Low-power is an important specification of SoC design and evaluation, a new low power design scheme was proposed by using weighted datapath. Firstly, the algorithm uses program slicing technictue to extract RTL data path.Secondly, the weights(frequcncy of use) of datapath arc obtained via Baycsian network training, and then the weighted datapath is generated. Finally, to reduce system power consumption effectively, the scheme controls the generation of clock gating logic, and it gives high priority for the datapath with low weight to insert or merge clock gating logic. Experimental results show that the proposed scheme has low computation cost, and it has 8. 38 %lower power consumption and 6. 8% lower hardware arear overhead when compared with existing low power SoC design scheme.

Design and Implementation of the Prophet Speculative Multithreading System

LI Zhong,ZHAO Yin-liang,DU Yan-ning

Computer Science. 2011, 38 (2): 296-301.

Abstract

PDF(597KB) ( 725 )

RelatedCitation | Metrics

Speculative Multithreading(SpMT) improves the performance by means of exploiting speculative thread-level parallelism. In SpMT, Thread runs speculatively and SpMT execution model is used to detect and recover from misspeculation. hhis paper described Prophet, which is a hardware SpM’execution model implementation. Prophet execulion model was described in details in this paper, including thread state control mechanism and Prophet Multi-versioning Cache system. Multi versioning Cache system provides the speculative data buffering functionality and uses a snooping bus based cache coherence protocol to detect data dependence violation. This paper also presented the evaluation of Prophet excution model via Olden benchmark, which proves that Prophet system could archive significant speedup to non-numcnc programs.

MD5 Crack Method Based on Compute Unified Device Architecture

ZHANG Run-mei,WANG Xiao

Computer Science. 2011, 38 (2): 302-305.

Abstract

PDF(956KB) ( 965 )

RelatedCitation | Metrics

CUDA is intended to give full play to the advantages of ultra-high computing performance of GPU in data processing, scientific computing and other fields of general purpose. This paper studied MD5 crack method based on Compute Unified Device Architecture and carried on hybrid compilation by using VS-2005 and NVCC. The experiment runs CUDA programs on GcForce-9600GT graphics card and runs Standard-C programs on Quad-Corc CPU-Q6600.The results shows that, under the environment of high computational load and huge amounts of data, the computing speed of mid and low end graphics card is 30 to 50 times higher than that of high end CPU. CUDA gives full play to the advantages of GPU Streaming　Multiprocessors Array and greatly improves the efficiency of the parallel computation　programs.