ISSN 1000-1239 CN 11-1777/TP

Table of Content

15 August 2006, Volume 43 Issue 8
Research and Design of a Cluster Multimedia Storage System
Wan Jiguang and Xie Changsheng
2006, 43(8):  1311-1316. 
Asbtract ( 358 )   HTML ( 0)   PDF (602KB) ( 547 )  
Related Articles | Metrics
With the increase of multimedia data in the Internet and the high-speed development of cluster, a PC-clustered high performance cluster multimedia storage system is presented, which adopts high autonomy and maintenance by itself to improve the parallelism and the expandability of the system; the CMSS adopts logic namespace management to realize a simple namespace and avoid single point of failure; it adopts data backup technology among the servers to improve the data security of the system and resolve the balance of load among the servers. In this paper the design of the CMSS is introduced in detail, and an intensive performance evaluation study on the CMSS is performed.
A Multi-Agent Cooperation Model Based on Dynamic Description Logic
Luo Jiewen, Shi Zhongzhi, Wang Maoguang, and Lin Fen,
2006, 43(8):  1317-1322. 
Asbtract ( 489 )   HTML ( 0)   PDF (592KB) ( 682 )  
Related Articles | Metrics
Dynamic description logic is proposed for formalizing multi-agent cooperation process with a clearly defined syntax and semantics. By combining the features of knowledge representation and reasoning of description logic and action theory for multi-agent interaction, the present logic is effective and significant for both static and dynamic environment. On the static side, a description logic is employed for the representation and reasoning of belief and goal. On the dynamic side, an object-oriented method is adopted to describe actions. The description of each action is composed of the models, preconditions and effects, which reflects the real changes of the world and is very suitable for belief revision and multi-agent planning. Based on the logic, how to form joint goal for multi-agent cooperation is investigated. In particular, an effective dynamic planning algorithm is proposed for scheduling sub goals, which is greatly crucial for coordinating multi-agent behaviors.
A New Data Clustering Algorithm for Parallel Whole-Genome Shotgun Sequence Assembly
Lin Jiao, Chen Wenguang, Li Qiang, Zheng Weimin, and Zhang Yimin
2006, 43(8):  1323-1329. 
Asbtract ( 462 )   HTML ( 2)   PDF (780KB) ( 837 )  
Related Articles | Metrics
Presented in this paper is a data clustering method based on graph-partition in parallel whole-genome sequence assembly. The algorithm transforms the data clustering problem into graph partition problem which helps to solve the load unbalancing in the parallel assembly stage. In addition, the method improves the quality of clustering by adding paired mate information into the read-relation graph which shows relationship between reads accurately. Experiments in both artificial and real genome data sets show that the data clustering method can obtain high quality clustered data and outperforms the traditional method significantly.
A Multiple Sequence Alignment Algorithm Based on a Hidden Markov Model and Immune Particle Swarm Optimization
Ge Hongwei and Liang Yanchun
2006, 43(8):  1330-1336. 
Asbtract ( 547 )   HTML ( 2)   PDF (729KB) ( 704 )  
Related Articles | Metrics
Multiple sequence alignment (MSA) is a fundamental and challenging problem in the analysis of biologic sequences. In this paper, an immune particle swarm optimization (IPSO) is presented, which is based on the models of the vaccination and the receptor editing in immune systems. The proposed algorithm is used to train hidden Markov models (HMM). Furthermore, an integration algorithm based on the HMM and IPSO for the MSA is constructed. The approach is examined by using a set of standard instances taken from the benchmark alignment database, BAliBASE. Numerical simulation results are compared with those obtained by using the Baum-Welch training algorithm. The result of the comparisons show that the proposed algorithm not only improves the alignment abilities, but also reduces the time cost.
UPIM: A User-Centered Pen-Based Interactive System
Wang Xiaochun, Tian Feng, Qin Yanyan, and Dai Guozhong
2006, 43(8):  1337-1344. 
Asbtract ( 589 )   HTML ( 2)   PDF (715KB) ( 1074 )  
Related Articles | Metrics
Pen-based user interface (PUI) is a primary style in the post-WIMP world, which has many potential advantages such as naturalness and being easy to use and to learn. However, the application-centered software design method cannot ensure the pen-based user interface to keep these advantages when pen-based interactive system is developed. Hence, a software design communication model is first analyzed. Then, based on this model, a user-centered software design method is presented, which is described in detail through the scientific diving training management system. The evaluations of the practical systems show that the method can effectively improve the usability of pen-based interactive systems.
An Approach to Monitoring and Controlling Workflow Systems Based on the Instance State
Yu Wanjun, Liu Dayou, Liu Quan, and Yang Bo
2006, 43(8):  1345-1353. 
Asbtract ( 352 )   HTML ( 0)   PDF (1010KB) ( 700 )  
Related Articles | Metrics
The monitoring and controlling of workflow process is an efficient method for dealing with the exception of workflow in run time and guaranteeing the enaction of workflow instances in the way of correctness and effectiveness. The monitoring and controlling of workflow based on states not only can cope with the exceptions from progression of workflow instances at real time, eliminate the retardance of instance enacting, but also make necessary intervention for instance running. But these operations often result in inconsistence of workflow model and workflow instances. It is quite often that inconsistency causes the workflow instance to be terminated if no corresponding adjustments could be taken. Firstly, a number of necessary constraint rules for model correctness and instance consistence are given based a workflow model—ADEPT, and then a states-based approach to monitoring and controlling is proposed, that includes a set of primitive transforming task instance from current state to next and algorithms transferring process instance from current status to future. The use of the approach guarantees both model correctness and instance consistence. Finally, completeness and consistence of the set of primitive are discussed briefly.
An Integrated Framework for J2EE-Based E-Learning Systems and Its Application
Li Wei, Luo Junzhou, and Cao Jiuxin
2006, 43(8):  1354-1360. 
Asbtract ( 289 )   HTML ( 1)   PDF (640KB) ( 583 )  
Related Articles | Metrics
With the popularization of e-learning, various teaching software, tools and e-learning platforms unceasingly appear. In order to solve the problems faced while integrating the heterogeneous e-learning systems, such as the heterogeneity, the interoperation and the scalability, an integrated framework for e-learning systems based on J2EE (IFESJ) is proposed. IFESJ peels off the common services from the heterogeneous e-learning systems and integrates them to support the teaching activities based on the network from aspects of the portal, the service and the data. The practical application and the testing results on the performance indicate that the design of IFESJ is reasonable and the e-learning system based on IFESJ can satisfy the concurrency requirements in practice.
An Efficient XML Documents Classification Method Based on Structure and Keywords Frequency
Yuan Jiazheng, Xu De, and Bao Hong
2006, 43(8):  1361-1367. 
Asbtract ( 409 )   HTML ( 1)   PDF (688KB) ( 751 )  
Related Articles | Metrics
According to the XML Web page character, an efficient method for computing XML document similarity, position weight and frequency of keywords in documents is presented. Then some features are selected from XML documents based on the method and a multi-classification algorithm of XML Web page is proposed using support vector machines. In this algorithm, a CFK(classifier feature kernel) of common similarity features is created from each sample set of XML documents class. The class label of an XML document is determined by computing similar distance between a test XML document and each CFK. Experimental results prove the effectiveness of the classification algorithm and good performance for multi-classification of XML documents.
Semantic Document Reference Metadata Extraction in SemreX
Guo Zhixin, Jin Hai, and Chen Hanhua
2006, 43(8):  1368-1374. 
Asbtract ( 421 )   HTML ( 2)   PDF (939KB) ( 788 )  
Related Articles | Metrics
In order to implement knowledge sharing between technical researchers, a new method using techniques of semantic Web is proposed, which can retrieve reference metadata from varied documents. By way of pattern matching, it can get some metadata such as authors, titles, publishing date and journal name, and uses the OWL ontology description language to formulate the metadata, which assists the semantic searching in the next step. Experiments prove its efficiency.
Research Progress in Testing Techniques of Component-Based Software
Mao Chengying and Lu Yansheng
2006, 43(8):  1375-1382. 
Asbtract ( 410 )   HTML ( 1)   PDF (803KB) ( 999 )  
Related Articles | Metrics
Software component technology provides a more effective design pattern than object-oriented methodology. Component-based software has been widely used in many fields and become a fairly popular software form in recent years. However, the features of component, such as information encapsulating and high evolvability, and properties of component composition (e.g., heterogeneous, loosely coupled and interoperability, etc.) bring great challenges to the testing of component-based software systems. In this paper, the representative testing methods and techniques for component and component-based software in recent years are analyzed and surveyed. Some effective testing frameworks and tools are also summarized and compared. Furthermore, the directions of future research are explored.
Self-Organizing Peer-to-Peer Network Based on Community Mimicking Social Society
Tang Jiuyang, Zhang Weiming, Xiao Weidong, and Tang Daquan
2006, 43(8):  1383-1390. 
Asbtract ( 400 )   HTML ( 1)   PDF (830KB) ( 678 )  
Related Articles | Metrics
Most of the existing peer-to-peer (P2P) systems assume that the peers in such a system form groups, yet little research has been done discussing how such groups form and remain stable as nodes interests change. Referring to the human society organization, a method of self-organizing P2P network based on community is proposed. According to the peer's resources and query requirements, the P2P network is organized by resource community and demand-oriented community. As resources and demands change, peers dynamically adjust their configuration and move to the right community gradually. The simulation result indicates that the proposed method can automatically optimize the network so that each peer can maintain near neighborship with peers with similar resources or provided resources. Also, the experiment proves that the employment of self-organization mechanism leads to adaptation as well as scalable P2P network construction technique.
A Centralized Warrant Distributing and Updating Approach for Secure Multicast Group Control
Zhao Xin, Li Xiaojian, and Wu Wei
2006, 43(8):  1391-1397. 
Asbtract ( 427 )   HTML ( 1)   PDF (931KB) ( 546 )  
Related Articles | Metrics
In order to control the communication relation actively in secure multicast, an approach is proposed, which distributes warrant containing secret key. Users get warrant in order to communicate in some multicast group. The group controller can update the users' warrants to change the users' communicate relation. This approach also uses an encryption algorithms and types' mapping table to increase the security. The result shows that this approach can set several smaller secure groups in a multicast group and change the users' communication relation actively, so as to act against attack or control the communication relation.
Triple-Block-Chaining-Based Hash Function and Its Performance Analysis
Huang Yuhua, Hu Aiqun, and Wang Xingjian
2006, 43(8):  1398-1404. 
Asbtract ( 476 )   HTML ( 0)   PDF (728KB) ( 561 )  
Related Articles | Metrics
A hash function based on triple-block chaining (HTBC) is put forward and its security is demonstrated. The speed of the HTBC algorithm is faster than that of hash functions in common use (SHA and MD family). The dependence test results accord with the demands. The HTBC algorithm is complete; its degree of avalanche effect is about 0.9993; its degree of strict avalanche criterion is 0.992 or so. The frequency test results indicate that the output generated by the HTBC algorithm has uniformity. The binary matrix rank test results indicate that it is linear independent among disjoint sub-matrices of the output. Maurer's universal statistical test results show that the output could be significantly compressed without loss of information. The results of run test, spectral test, non-overlapping template matching test, overlapping template matching test, Lempel-Ziv compression test, linear complexity test, serial test, approximate entropy test, cumulative sums test, random excursions test, and random excursions variant test all fulfill the requirements. Therefore the output generated by the HTBC algorithm has good pseudo-randomness. Thus the security of the HTBC algorithm is verified by way of statistical evaluation.
A Multi-Candidate Electronic Voting Scheme Based on Secure Sum Protocol
Zhong Hong, Huang Liusheng, and Luo Yonglong,
2006, 43(8):  1405-1410. 
Asbtract ( 479 )   HTML ( 2)   PDF (548KB) ( 1038 )  
Related Articles | Metrics
Multi-candidate electronic voting is very useful in numerous practical situations. Most of the previous schemes discuss Boolean vote in which voters can only cast “yes/no” vote. A novel multi-candidates election scheme is presented which is appropriate for k-out-of-m election. The main idea is to express a ballot by a multi-precision number that allows voting for up to k out of the m candidates in one ballot. The purpose of the vote is to elect more than one winner among m candidates. The self-tallying protocol in the solution is based on the multi-precision arithmetic and the secure sum protocol without any trusted third party. In comparison with general methods, the results of this paper don't rely on any traditional cryptography or computation intractability assumption that is information-theoretically secure. This scheme achieves security including perfect ballot secrecy, receipt-free and dispute-freeness tally that extend the previous general security. Each participant takes only O(nm(log_2 n)) bit operations, which is more efficient than previous schemes. The scheme is very practical and can be efficiently implemented.
The PBL Method: A Novel Parallel Error Detection Method for Intrusion Tolerance Systems
Li Qinghua and Zhao Feng
2006, 43(8):  1411-1416. 
Asbtract ( 352 )   HTML ( 2)   PDF (685KB) ( 566 )  
Related Articles | Metrics
One of the most advanced research problems in intrusion tolerance systems (ITS) is error detection, which has become another essential technique in system security to prevent the intrusion from generating a system failure. A parallel error detection method named PBL for ITS, which is based on distributed Bayesian learning, is proposed in this paper. This method is particularly suitable for detecting errors with distributed sources. The PBL method not only is useful in detecting errors in the distributed network environment, but also can be used to enhance noise tolerant ability of ITS. Some key problems of the PBL method in detail are also discussed.
Algorithm Research on “On the Fly” Model Checking Temporal Logics of Knowledge in Multi-Agent Systems
Wu Lijun, Su Kaile, Chen Qingliang, and Yang Zhihua
2006, 43(8):  1417-1424. 
Asbtract ( 517 )   HTML ( 0)   PDF (837KB) ( 589 )  
Related Articles | Metrics
Temporal logics of knowledge have been widely used in the distributed systems community and in the expression for the specifications of protocols. The model checking for temporal logics of knowledge becomes a new and important research domain. So approaches to the “on the fly” model checking the temporal logics of knowledge are discussed. Based on the “on the fly” model checking approaches for temporal logics, and according to automata theory and the semantics of knowledge, the “on the fly” model checking approaches to the temporal logics of knowledge are presented. These approaches, making the model checking for the specifications with knowledge operators, need only to construct a small portion of state space of system before a counterexample is found, and so avoid memory-shortage and state-explosion and realize “on the fly” model checking for the temporal logic of knowledge. And the time complexity of the algorithm is polynomial. Finally, the application to the verification of the TMN cryptographic protocol is illustrated to show the effectiveness of the approach.
The Directional Similarity-Based Clustering Method DSCM
Xiu Yu,, Wang Shitong,, Wu Xisheng, and Hu Dewen
2006, 43(8):  1425-1431. 
Asbtract ( 484 )   HTML ( 3)   PDF (789KB) ( 765 )  
Related Articles | Metrics
The directional similarity-based robust clustering approach DSCM for directional data is presented in this paper. The DSCM utilizes the objective function based on the directional similarity measure, and then optimizes it using the fixed point iteration method such that all the stable states of the data samples are derived. By presenting all these stable states to the hierarchical clustering algorithm AHC, the final clustering results are obtained. This new approach exhibits its robustness to initialization and capability to reasonably detect the number of clusters for clustering directional data. The experimental results demonstrate its validity and distinctive superiority over the two conventional directional clustering algorithms.
Application Research of Cognitive Physics in Data Mining
Yang Bingru, Gao Jing, and Song Wei
2006, 43(8):  1432-1438. 
Asbtract ( 409 )   HTML ( 3)   PDF (684KB) ( 761 )  
Related Articles | Metrics
Cognitive physics, from the perspective of natural language, focuses on the thinking process that is from quantity to quality and from data to knowledge as well as the organization of information. In this paper, data mining techniques are explored by using cognitive physics theory and its relevant theory. Firstly, by using the theory of field, the language field is illustrated, which is a method for representing complex knowledge. Secondly, by learning from the information spreading theory, the parameter evolution is discussed, which is an important issue for post-processing of data mining. Finally, by analyzing the information entropy method for decision tree construction, the subjective entropy and the corresponding SID\-3 algorithm are proposed. The advantage of the proposed algorithm is illustrated by using an example.
Interaction-Oriented Ability Specification and Reasoning
Zhang Hui, and Li Sikun
2006, 43(8):  1439-1444. 
Asbtract ( 358 )   HTML ( 0)   PDF (550KB) ( 501 )  
Related Articles | Metrics
The interaction among different agents is an important factor of the agent capabilities in multi agent systems. Presented in this paper is a scheme that can describe the impact of interaction on agent capability by integrating description logic into task languages. A formal system DTL is proposed, which is intended to axiomatize the set of accomplishable task formulas. The soundness, completeness and decidability of DTL are also proved.
Long-Term MAS Coalition Based on Fuzzy Relation
Tong Xiangrong and Zhang Wei
2006, 43(8):  1445-1449. 
Asbtract ( 397 )   HTML ( 3)   PDF (544KB) ( 612 )  
Related Articles | Metrics
Long-term coalition of MAS has several advantages such as decreasing cost, maintaining coalition stability and improving relation. Breban and Griffiths researched trust relation among agents. But there is no feasible relation algorithm. A fuzzy relation of long-term coalition is proposed which can compute the relation value of members easily and effectively. This method reasonably describes fuzzy relation of long-term coalition. Through monitoring agent interaction, it can update the fuzzy relation. Its advantages are the decrease of computing complexity and the effectiveness improvement. It improves the work of Breban and Griffiths. The simulation result demonstrates the reliability and practicality of this method. The method is the basis of forming and evolvement research of long-term coalition and organization.
GLTO Design and Analysis
Chen Yu, Zhu Xiaojing, Zou Qiong, and Liu Ling
2006, 43(8):  1450-1456. 
Asbtract ( 390 )   HTML ( 0)   PDF (1088KB) ( 519 )  
Related Articles | Metrics
Link time optimization is a technique which optimizes the whole program after compilation. It overcomes the limitations of traditional compilers by enlarging the optimizing scope from a single function or a module to the whole program, and fully utilizes the information only available at link time. Guided by the link time optimizer ALTO developed by Arizona University for the Compag Alpha, the GLTO (Godson link time optimizer) is designed with the consideration of the features of the microarchitecture and the instruction set of the Godson-2 processor. GLTO can achieve a 9.4% performance improvement of the SPEC2000 integer benchmark with ref input on average. In this paper, the effect and the cause of several major optimizing methods are explored, some improvements in the processor architecture design are proposed, and the differences between GLTO and ALTO are analyzed.
A Study of Software Test Criterion Effectiveness Measure
Zhao Liang, Wang Jianmin, and Sun Jiaguang
2006, 43(8):  1457-1463. 
Asbtract ( 555 )   HTML ( 3)   PDF (795KB) ( 1077 )  
Related Articles | Metrics
How to measure the effectiveness of test criterion is an important problem for software quality assurance. Till now, no consistent measures have been achieved. In this paper, potential failure distance (pfd) is proposed as an indication of the effectiveness of test criterion. It regards the effectiveness of test criterion on particular software as the function of specification, program, fault type and criterion. It proceeds measure from the point of interaction between software and test criterion. Compared with other related work, it reflects the achieved confidence on the correctness or reliability about the software after testing. The measure provides a feasible comparison of the effectiveness of test criteria on software. The measure can help to design more testable software.
Concurrent Behavior Control for Active Rules of Spatial Database
Xiong Wei, Liao Wei, Chen Hongsheng, and Jing Ning
2006, 43(8):  1464-1470. 
Asbtract ( 330 )   HTML ( 2)   PDF (759KB) ( 511 )  
Related Articles | Metrics
There are usually a great deal of rules triggered simultaneously in complicated spatial information applications in network environment. Therefore the efficiency of rules concurrency is more important. Most methods focus on maintaining accuracy of database status and ensuring the performance of simple concurrent operation. Therefore solution to control concurrent rule set are studied. Labeled events graph and termination analysis algorithm are presented, which are illustrated by an example. A hybrid indexing method, the HRlink-tree, is proposed, based on which an improved bottom-up update algorithm is presented. Simulation experiment shows that the HRlink-tree outperforms the Rlink-tree in update performance.
A Register Allocation Algorithm for Predicated Code
Wang Fengqin, Hu Dinglei, and Liu Chunlin
2006, 43(8):  1471-1476. 
Asbtract ( 507 )   HTML ( 1)   PDF (646KB) ( 811 )  
Related Articles | Metrics
In order to allocate registers efficiently for predicated code, a predicate analysis system based on binary decision diagrams is introduced first, which is put forward by John W. Sias et al. And then the traditional register allocation procedure is improved, and a new algorithm for constructing refined interference graph is presented. The algorithm have been implemented in the compiler of YHFT-DSP/700 chip developed by the authors' college. Experiment results show that the number of used registers is reduced, the code execution time is shortened, and the performance is improved greatly.
A Novel Method for Facial Pose Discrimination and Frontal View Synthesis
Chen Jiada, Lai Jianhuang, and Feng Guocan
2006, 43(8):  1477-1484. 
Asbtract ( 519 )   HTML ( 1)   PDF (1015KB) ( 817 )  
Related Articles | Metrics
Facial pose under depth rotation is a tough problem in human face recognition. In this paper, it is firstly discussed that orientation and size of Gabor filter have the optimal capability to discriminate face pose. Secondly, a Gabor-based KFDA(kernel Fisher discriminant analysis) method coupled with fractional power polynomial model is proposed to serve as the face pose classifier. Finally, an integrity human face synthesis method is presented based on a new full-correspondence algorithm and linear object classes. The experiment results show that the proposed method is efficient.
Application of Dimension Reduction on Using Improved LLE Based on Clustering
Wang Heyong, Zheng Jie, Yao Zheng'an, and Li Lei
2006, 43(8):  1485-1490. 
Asbtract ( 543 )   HTML ( 2)   PDF (700KB) ( 977 )  
Related Articles | Metrics
Locally linear embedding (LLE) is one of the methods intended for dimension reduction. Its extension using clustering and improved LLE for dimension reduction is investigated. Firstly, using clustering can reduce time-consuming. Secondly, the improved LLE is suitable for selecting the number K of the nearest neighbors. When the number K of the nearest neighbors is small, it can obtain good results. While the original LLE algorithm obtains the same results, the number K of nearest neighbors may be much larger. Even if the number K of the nearest neighbors using the improved LLE is selected to be larger, the result is still right. So, the improved LLE is not sensitive to the selection of K. It is shown that the improved LLE based on clustering has less computing than the original LLE algorithm and enlarges the choice of parameter K by experiment.