Abstract:Nowadays, workflow technology has been intensively applied in the domains of administration, production, and scientific research. For the complexity and variability of process logic, how to build a well-structured business process is a practical problem. Thus, process model analysis and improvement become one of research directions in workflow technology. Firstly, according to the modified workflow process meta-model, a P/T system based workflow model procedure net is proposed. And then, a set of simplification rules are given to verify two structural conflicts of the process model, i.e. deadlock and lack of synchronization. Finally, following these rules, a procedure net example transformed from a business process is simplified, and the result is obtained.

Abstract:Emerging service oriented architecture is increasing the evolution and variation of software resources. Applications that cannot adapt to dynamic environments will decrease their usefulness, particularly to business process management systems that face requirements changed frequently. In response to the realistic requirements, this paper, based on Web services and business process management techniques, propose a model of service cooperation middleware (SCM) for enterprise computing. First, it discusses the conceptual architecture and operation mechanisms of SCM, then formalizes the meta-model of cooperative processes deployed on SCM. By introducing a model transformation function into the formalization, a static process model can be extended to a dynamic one. In virtue of the reflection capability, structural reconstruction and behavioral adaptation of the dynamic processes can be achieved in SCM by introspection and effectuation. In terms of the conceptual model of SCM, a process virtual machine (PVM) is designed as a running container for cooperative processes. Borrowed from the power of MDA mechanism, the model of a business process can be transformed successively from design to run time. The business processes established on SCM can be more flexibly and extensively applied to various open environments, and leverage modern enterprise computing.

Abstract:Simulation-Based verification approaches need a large amount of test vectors for verifying the corner case of designs. This paper proposes a novel assertion-based automatic functional vectors generation method which takes assertion as the functional coverage metric. For the given assertion, the related design part is first extracted, and then the assertion property and the signal propagation process based on the DD model is converted to the CLP constraints; Finally, functional vectors are generated by solving the constraints. The advantage of this method is the combination of the program slicing based design extraction, the word-level SAT engine, and the dynamic search techniques. The method can deal with large designs and handle control and datapath design in a unified framework. Experimental results show that the method is efficient for finding the design errors and vectors generation.

Abstract:Recently, many research activities have targeted at reducing energy consumption of computer systems through software optimizations. By using the functions of dynamic voltage scaling (DVS) and turning off unused system units (TOSU), some software techniques can be used to reduce the wasted energy consumption, and one of the techniques is the compiler-directed method. Many hardware details are complex and fuzzy about DVS and TOSU, and so the paper has established an analytic model for compilation. Based on analyzing the model, the paper gives the idea about localizing the use of system units. If the systems have functions of DVS and TOSU, localizing the use of system units is an effective method of low-power compilation optimization.

Abstract:IPSM is an integrated probabilistic image semantic description multi-level model. This model includes input layer, feature layer, semantic layer, synthetical probability layer, probability propagation layer, and semantic mapping layer. Based on the model and characterizing of the image high-level semantic content according to Bayesian theory, SHM (semantic high-level retrieval algorithm) and SRF (high-level semantic relevance feedback) for image retrieval based on high-level semantic content, for user relevance feedback respectively, are designed and implemented. Experimental results indicate that IPSM, SHM and SRF are effective in characterizing image high-level semantic content and can provide sound and robust image retrieval performance.

Abstract:Fuzzy objective information systems (FOISs) exist in many applications, and knowledge reduction in them can抰 be implemented by reduction methods in Pawlak information systems. This paper firstly presents new reduction methods including a distribution, a maximum distribution, a assignment, and rough distribution reductions. It then probes into their properties and the relation between them and the reduction methods on Pawlak information systems. Furthermore, this paper proposes the judgement theorems and discernibility matrixes with respect to these reductions. These reductions extend the corresponding methods in Pawlak information systems and provide a new and low computation complexity way for knowledge discovery and rough-fuzzy rule based fuzzy concept classifiers in fuzzy objective information systems.

Abstract:Current neural network ensemble methods usually generate accurate and diverse component networks by disturbing the training data, and therefore achieve strong generalization ability. In this paper, causal discovery is employed to discover the ancestor attributes of the class attribute on the results of the sampling process. Then, component neural networks are trained on the samples with only the ancestor attributes being used as inputs. Thus, the mechanism of disturbing the training data and the input attribute is combined to help generate accurate and diverse component networks. Experiments show that the generalization ability of the proposed method is better than or comparable to that of the ensembles generated by some prevailing methods.

Abstract:The authors investigate carefully whether or not two classes of inference algorithms, which are Compositional Rule of Inference (CRI) proposed by L.A.Zadeh and Triple-Implication algorithm (Triple-I) proposed lately, hold the continuity and approximation properties, and moreover, also how the approximation errors are propagated by them. Therefore, a fuzzy inference algorithm is viewed as a mapping from one fuzzy set to another, Hamming distance formula is used as the computing distance between the two fuzzy sets. The authors prove that the two classes of algorithms hold the continuity properties in the cases of fuzzy modus ponens and fuzzy modus tollens. The authors also point out that Triple-I algorithm always holds the approximation property if antecedent and consequence of the known rule are normal fuzzy sets. However CRI algorithm holds approximation property only if CRI holds consistency property. Two classes of algorithms do not make approximation errors magnified when they hold approximation property. The results of the paper are useful for the selection and analysis of algorithms for fuzzy inference when practical fuzzy control and expert systems are designed.

Abstract:Graphics processing unit (GPU) has been developing rapidly in recent years at a speed over Moor抯 law, and as a result, various applications associated with computer graphics advance greatly. At the same time, the highly processing power, parallelism and programmability available nowadays on the contemporary GPU provide an ideal platform on which the general-purpose computation could be made. Starting from an introduction to the development history and the architecture of GPU, the technical fundamentals of GPU are described in the paper. Then in the main part of the paper, the development of various applications on general purpose computation on GPU is introduced, and among those applications, fluid dynamics, algebraic computation, database operations, and spectrum analysis are introduced in detail. The experience of our work on fluid dynamics has been also given, and the development of software tools in this area is introduced. Finally, a conclusion is made, and the future development and the new challenge on both hardware and software in this subject are discussed.

Abstract:Real time collision detection is required by almost all computer graphics applications. However, the problem of real time collision detection is yet to be solved between complex objects. With the recent advent of high performance graphics processing units (GPUs), a dramatic shift is being seen in the traditional idea that general-purpose computation can only be performed by CPUs. This paper explores to solve the problem of real time collision detection between complex objects using programmable GPUs. The algorithm maps the computation of collision detection between two arbitrary objects onto programmable GPUs to match their parallel architectures, and produces on the fly the collision detection results via real time rendering. To do so, the problem is first converted into the problem of finding intersections between a collection of line segments and a set of triangles to realize the migration of collision detection algorithms to programmable GPUs. Based on reasonable analyses of the algorithm complexity, two optimized techniques is presented to improve its efficiency. Experimental results have shown that the optimized algorithm is advantageous over other current collision detection algorithms implemented in image space regarding efficiency, accuracy as well as practicability.

Abstract:In this paper, a novel hybrid multiresolution approach is proposed to efficiently render highly detailed objects based on the respective merits of point-based and polygon-based rendering. Both point and polygon are tightly integrated into the uniform structure of P-Quadtrees, which are constructed from geometry images. While traversing the P-Quadtrees in rendering, the part of surface that faces the viewer can be rendered by coarser quad mesh to reduce the numbers of rendering primitives. The shading details can be enhanced by hardware accelerated normal mapping. The view-dependent LOD (level of detail) selects the finer hierarchy on silhouette, which is rendered by points. The rendering efficiency of large-scale model is greatly improved while the visual effect at both the surfaces and the silhouette is guaranteed.

Abstract:By means of optimization methods, degree reduction of rational Bzier curves is changed to an optimization problem so that both weights and vertices are considered respectively. Using programming method and Genetic Algorithms, a new method on the reduction of rational Bzier curves is presented. The method has the following virtues: Firstly, it is simply to get the result by fitness function, copy process, crossover process, mutation process, and selection process. Secondly, the rational Bzier Curves can be reduced many times and interpolated. Finally, the reduced Bzier curves can be represented explicitly.

Abstract:Information stored in XML documents should be protected by access control policy. Current access control models for XML documents are all based on DAC (discretionary access control) or RBAC (role-based access control). High security system uses MAC (mandatory access control) to secure information in system. XML document model is extended to include label information in this paper, and some rules that the extended model has to satisfy with are presented. Fine-grained MAC model for XML documents is described in detail by discussing four operations on XML documents. The fine-grained MAC model is based on XML schema, and its finest granularity of access control is element or attribute. The architecture and some mechanisms used to implement the fine-grained MAC model are discussed too.

Abstract:PGM (pragmatic general multicast) is a reliable multicast transport protocol, which runs on IP networks. Because PGM standard does not include any congestion control scheme in itself, it can’t achieve real-time response to satisfy the requirements of network and can’t timely adjust the sending rate. Based on the PGM protocol, a novel close-loop controller, which is suggested to be implemented between the sender and the CLR (current limiting receiver), is introduced in this paper to timely adjust the sending rate. The proposed rate control scheme brings the source’s sending rate of the network to be stable gradually and has fast response. Simulations are carried out for the proposed control scheme, and the simulation results demonstrate that the proposed algorithm has the following properties: scalability, stability, fast convergence, and fast response under various network environments. They show good performance of the proposed multicast control schemes.

Abstract:This paper presents a model which can support network security objects, improve the Amon ott’s rules with small amount of operations and storages for practicality, enhance the flexibility available for system implementation by making the single level becoming level range , and control IPC objects effectively. For these purposes, the Amon ott’s rules for dynamically modifying the current sensitivity level are extended to ones for sensitivity levels range, so Bell's work on making the single level becoming level range for network security can be combined with Amon ott’s. Considering the cases in the practical system GEMSOS,DG/UX and prototype microkernel system Fluke, single level entity, multiple level entity and special access mode for progress, and the invariants and constraints corresponding to them are introduced. Based on Tmack’s way, a sufficient mechanism for IPC objects is posed. In addition, some flaws in ABLP model are pointed out. A new confidentiality policy model with formal specification of invariants, constraints, variables, and constants has been presented with demonstrating reasonableness for some constraints, and it can be used for system design.

Abstract:Arnold transformation is applied widely in digital image encryption because of its periodicity. But if its periodicity is used in the course of image decryption, it will waste lots of time. So this paper offers a new anti- Arnold transformation algorithm. The algorithm gets anti-Arnold transformation by solving equation groups. The application of Arnold transformation from square image to rectangle image is generalized and its application on image encryption expanded. At the same time, the anti-Arnold transformation from 2-dimension to m-dimension is also generalized.

Abstract:Service discovery is an important aspect of network self-configuration. Solutions to local area service discovery had been put forward and applied to practical systems for several years. In terms of wide area network, the scalability problem of service discovery system must be addressed because there are numberless dynamic services and clients distributed in wide area. To get good scalability, service discovery system should maintain right function and steady performance when the number of services and clients increases continuously. A hash-value based scheme for wide area service discovery is presented in this paper. Based on the model of service description, service query, and match condition, the sets of type name and attributes name in service description and service query are mapped into hash-value and hash-value area respectively. They are used to identify the service directory that will record the service description or satisfy the service query. Each service directory in wide area network is assigned a value area to indicate the range of service descriptions to be recorded in it. All of the service directories are organized into a content addressable tree according to the relationship of their value areas. In such systems, service description is forwarded to certain directory to be recorded in terms of its hash-value, and service query is forwarded to the directories that might satisfy it in terms of its hash-value, area. Simulation results, along with analysis and comparison, show that this scheme can avoid deficiency of the existing schemes and gain favourable scalability.

Abstract:In this paper, review results of the General Programs ended in December 2003 funded by Computer Science Division of Information Science Department of NSFC are presented. Meanwhile review standards are also presented for the reference of applicants and undertakers.