Abstract:

Abstract:XYZ system is a programming support system with the goal to enhance reliability and productivity of software development. It consists of a TLL (temporal logic l anguage) XYZ/E to serve as its kernel and a suite of software engineering tools. In order to study the application of XYZ system to the multimedia field, in thi s paper, the authors present a method for automatically generating synchronizer in XYZ/RE from the temporal specification of a multimedia program. XYZ/RE is a s ub-language representing real time system in XYZ/E. Compared with the related w ork, this method can transform not only a simple temporal specification but also a nested temporal specification, so that the generated synchronizer can be reus ed for different purposes.

Abstract:At run-time, some parameter values may cause d uplicate function calls which cannot be found at compile-time by traditional op timizer. In this paper, the authors propose a run-time optimization method. It can eliminate the duplicate function calls by creating function call dependency graph.

Abstract:In this paper, the authors propose a new approach to requirements analysis, call ed OORA (ontology-oriented requirements analysis). The most important feat ure of this approach is that various ontologies have been introduced to represen t relationships between objects and/or classes. These ontologies may significant ly enhance the expressiveness of the current OOA (object-oriented analysis) app roaches. In this paper, first the OORA requirements model is designed. After tha t, two key types of ontologies are illustrated by two typical use cases. Finally , the main phases in OORA are given.

Abstract:With the development of mobile agent technology and the widespread of component ware, the requirements for a new componentware framework based on mobile age nt technology are increasing. After analyzing the limitations of traditional com ponentware framework, the authors propose a new componentware framework in this paper, which has better dynamic adaptability to network environments than the tr aditional componentware framework.

Abstract:In this paper, the authors describe a visual language XYZ/ADL (XYZ/architecture description language) for software architecture description based on XYZ/E to su pport a new approach of architecture-based programming. XYZ/ADL is composed of components, connectors and ports, by which it can describe many commonly used ki nds of software architectures. Since XYZ/E can represent both statistic and dyna mic semantics in a unified form, XYZ/ADL which is built on it can finish smoothl y the transformation between architectures at different abstract levels in a uni form frame.

Abstract:In this paper, a method is presented to partition productions of CFG (context-f ree grammar). It divides production set into two parts. The derivation with prod uctions in one part will never terminate, while it must terminate rapidly with p roductions in the other part. It is proved that the procedure of generating sent ences of CFL (context-free language) is using productions in one part to make t he sentential form longer and more complex first, and then using productions in the other part to terminate the procedure. A general controllable method is atta ined for generating sentences of CFL with restricted length or depth. The time a nd space complexity for generating one sentence is O(r+n), where n is th e restricted length or depth of sentences and r is the number of productions in given CFG. The generating strategies for different conditions are also discu ssed.

Abstract:Because of concurrency and non-determinism, the inner-states of a running dist ributed program should be considered a part from the start and the end states, w hen the program is to be tested and the validated. According to specification ba sed testing, these inner-states will tranform into event sequences through the ports of the program, and the co-relations among these events, i.e. the set of event sequencing constraints should be provided by the program's specification. In this paper, the authors introduce the E-CSPE (extended-constraints on su cceeding and preceding events), a method to formalize these event constraints. T he E-CSPE includes three basic discription rules, which correspond to three dif ferent types of such event constraints. Based on the consistency with and covera ge of these event constraints by the event sequences produced by the program exe cution, the validity of the distributed implementation can be tested.

Abstract:The advantage of COOZ (complete object-oriented Z) is to specify large scale so ftware, but it does not support refinement calculus. Thus its application is con fined and it can not be taken as a complete method for software development. I ncluding refinement calculus into COOZ remedies its disadvantage during design and implementation. The separation between the design and implementation for st ructure and notation is removed as well. Then the software can be developed smoo thly in the same frame. In this paper, development model is established, which i s based on COOZ and refinement calculus. Data refinement and operation refinemen t are debated with a example. As for implementary technology of refinement calcu lus, a data refinement calculator is constructed and an approach for data refi nement which is based on data refinement calculus and program window inference is provided.

Abstract:Connector is an important concept in software architecture. Since an engine mech anism is introduced into connector, connectors become triggers when they are con nected with components, and can be transformed into visible parts in implementin g system. This kind of connector is called active connector. In this paper, the authors introduce the concept of the software architecture that is based on acti ve connectors, and present a description language.

Abstract:This paper puts forward a test system framework TFDS (test system framework for distributed software system) for the testing of distributed programs, and as an illustration, introduces its prototype implementation PSET* (distributed progr am structure and event trace, revised version) on heterogeneous network platform s. The main functions of TFDS are for unit testing and integration testing. And the framework is divided into two parts: one for specification design and source code analysis, which works statically; and the other for program execution and event sequence manipulation, which works dynamically. Along with TFDS, and by a component-based architecture, the functionality of the PSET* can be easily im proved and reinforced.

Abstract:Mobile agent technology can be used as one of the key technologies for new compo nentware frameworks. The communication mechanism of mobile agent plays a very im portant role in mobile agent systems. Based on the design of the mobile agent sy stem——Mogent, after analyzing the mobility's influence on its communication me chanism, the authors propose a systematic solution to communication mechanism us ed in Mogent system in this paper. This mechanism involves agent naming, agent a ddressing and agent communication failure solutions, and its high availability, performance and robustness are demonstrated.

Abstract:In this paper, the authors present a complete introduction of a program developm ent and verification system based on rewriting techniques, focusing on the theor y, methods and techniques of the verification subsystem. The verification subsys tem enables the system to prove the correctness of the optimization rules and te st equations in programs and specifications, hence the soundness of the program development process is further guaranteed. The main technique employed in the ve rification subsystem is rewriting induction featuring batch proof method and wit nessed test sets.

Abstract:Data refinement is hard to deal with in a refinement tool compared with ordinary algorithmic refinement, since data refinement usually has to be done on a large program component at once. So it is important to learn how to perform data refinement effectively. In this paper, the background is introduced first, then the data refinement calculator is constructed. Finally an approach is proposed for data refinement which is based on data refinement calculus and program window inference.

Abstract:Development based on software architecture is one of the most promising methods for improving software quality, reducing software cost and raising software prod uctivity. Viewing architecture elements as first-class entities, the authors pr esent a specification method based on framework and role type to alleviate the p roblems that conventional methods may bring, and define a specification language FRADL (framework and role-type based architecture description language) in thi s paper. In FRADL, framework is considered as a component and role type with act ive connecting faculty is defined as a connector. The instances of components an d connectors are configured to compose the architecture for a family of software systems.

Abstract:An algorithm for volume rendering in an environment of parallel virtual machines is presented in this paper. In order to reduce the communication cost, as well as to guarantee the locality of all subtasks, the volume data are divided and or ganized as a series of slices. By maintaining and employing a database of the pe rformance index, the task subdivision algorithm produces an acceptable load bala ncing. An asynchronous binary method is introduced, which merges all partial ima ges in O(logn) time. An efficient developing platform based on the TCP/I P and Socket standards is built. It helps researchers to parallelize various ren dering algorithms on a virtual machine. The algorithm introduced in this paper h as been implemented on this platform, exploiting the classical client/server par adigm. The scalabilities of both the task size and the number of hosts are teste d. The experimental results are demonstrated and analyzed.

Abstract:In this paper, some shortages of traditional BP learning algorithm are analyzed. To avoid these shortages, a modified BP learning algorithm is proposed. It is s hown that this algorithm is super-linearly convergent under certain conditions. This algorithm can overcome some shortages of traditional BP learning algorithm , and has the same order of computation complexity as the traditional BP algorit hm. Finally, two computing examples are given. Simulation results illustrate tha t this algorithm is highly effective and practicable.

Abstract:Static task scheduling on network of workstations is well-known to be an NP-co mplete problem in a strong sense. Some heuristic algorithms have been proven to be sub-optimal under some restrictive conditions. In this paper, the authors pr esent a heuristic algorithm named DAG (directed acyclic graph) partition and sub -graph reconfiguration algorithm, which is a fast and effective one used in par allel task scheduling. The complexity of this algorithm is O(log^|V|I1518 ×(|V|+|E|)). It adopts recursion to implement DAG partition and sub-graph re configuration, then builds task clusters to carry out the task scheduling. At th e same time, it even optimizes the number of processors to some degree for it ha s not been solved before. The performance has been observed in a representative example compared with other existing scheduling schemes in terms of several valu able factors. The experimental results show that this algorithm is feasible.

Abstract:In this paper, the authors propose a method for mapping QoS (quality of service) guarantees between ISs (integrated services) and DSs (differentiated services) in the Internet environment. Based on the existing classification, this paper gi v es the mapping rules and relations between IS and DS, data flows and PHBs (per h op behaviors) according to different media streams. As the crux of mapping, mark ing is divided into two phases: pre-mark and mark. The pre-code marked in IS c onveys characteristics of individual flow into DS. According to this pre-code, boundary routers in DS remark packets and map it to relevant PHB. The authors al so propose a marking algorithm which is based on the TCA (traffic condition agre ement), taking account of the requirements of traffic control and fair link shar ing among different application traffics.

Abstract:In this paper, a new general binary tree drawing algorithm frame is de signed by using genetic algorithms. Under the frame, according to different appl ications, different binary drawing algorithms can be obtained by designing diffe rent objective functions reflecting aesthetic criteria. Furthermore, it is easy to add or reduce some of the aesthetic criteria in order to satisfy different us ers. Compared to previous algorithms, this algorithm is of the following advanta ges: the frames of the algorithms are unified, the method is simple, implementat ion and revision are easy. It has the following characters: self-adaptive, self -study and easily parellelized.

Abstract:The implemented BDI (belief, desire and intention) Agent systems used the three major mental attitudes as data structures, rather than as model operators. So th ere still remains a large gap between theory and practice. In this paper, the au thors present an improved programming language——AgentBDI, concerning the i nterpreter of dMARS (developed multi-agent reasoning system) after in-depth st udying PRS (procedure reasoning system) and dMARs, and provide AgentBDI's gramma r and operational semantic using a classical example of DAI. The authors also pr ovide an improved interpreter of BDI Agent based on AgentBDI. AgentBDI can b e viewed as an abstraction of dMARs, and allows BDI Agent programs to be written and interpreted in a manner similar to the logical programs.

Abstract:For speech recognition systems under noisy environment, lip-reading technique c an effectively reduce the influence of noise and improve the accurate rate o f speech recognition system by adding visual information to acoustic channel. In this paper, an effective and robust approach for lip and mouth locating and tra cking is presented to enable the information extraction under abnormal illumina tion and without special marks. This approach first locates face region with skin-color model, then finds the eyes from the face region with iterative algo rithm, modifies the position and size of face according to the position of eyes, transforms the lower part of face by specific color coordinators to clearly dis tinguish lip color from skin color, and finally describes the outline of upper l ip and lower lip with deformable template.

Abstract:During (partial) key escrow, monitoring is a vital phase. So how to monitor a us er safely and efficiently is a very important problem. According to the known mo nitoring schemes, after a user is monitored, his secret key is known. It is unfa ir to an honest user. In this paper, the authors propose a monitoring scheme of a typical partial key escrow scheme. In this scheme, the escrowed key of a user is not compromised even if the user has been monitored for many times.