Abstract:JIAJIA and MPI (message passing interface) are compared in this paper. JIAJIA and MPI represent the programming model of share memory and message passing respectively. MPI handles message transmission manually and is hard for programming. JIAJIA maintains data coherence by lower layer and provides simple message passing APIs, so it is easy and flexible to program. It takes more time for JIAJIA to initialize applications than MPI. A rough empirical formula about speedup and machine size is presented in this paper. Experimental results show that both JIAJIA and MPI have good parallel performance on PC clusters and their performance difference of most testing applications is lower than 10 percent. Applications with fewer messages have better performance and the gap between JIAJIA and MPI is minor. While messages increase, parallel performance decreases and the gap between JIAJIA and MPI depends on the amount of messages they produce.

Abstract:Based on the research of unordered tree-inclusion matching, a matching algorithm for XML-based component query is proposed. This algorithm can greatly improve the recall and provide support for Boolean query while maintaining a high level precision. Moreover, by adding some constraints on the basis of features of software component and using dynamic programming, the computation of matching cost is resolved in polynomial time, so that a high efficiency for the component query is guaranteed. Furthermore, the feasibility and efficiency of the new matching algorithm in practical application to software component query are confirmed by the results of a series of experiments on a prototype system RCRS.

Abstract:Applications with real-time constraints are not only growing in the field of embedded system, but gaining popularity in the desktop environment as well. In order to satisfy the need of users in these aspects, RFRTOS real-time OS based on Redflag Linux is developed. In this paper, the real-time support about real-time scheduling, preempted kernel, and fine-grain timer in RFRTOS is mainly discussed. The experimental results on RFRTOS show that the improvement gives a better result for applications with real time constraint than original Linux kernel.

Abstract:Component-Based development method is thought to be the latest outcome to tackle software crisis, but in practice it didn抰 reach the expectation. The methodology of component design is not perfect due to the lack of explicit guidance to build the practical application. In this paper, a business component design method is given, which uses the Cluster Algorithm to identify two kinds of business components (process component and entity component) and takes the requirement models as the data source of Cluster Algorithm samples. Several formulations are also given to calculate the value of samples?relationship. 揅ore Entity?was chosen to achieve better accuracy of Business Component identification, and granularity formulation is given to reduce the complexity of Business Component. The design steps of Business Component Interface are described at last.

Abstract:Architecture recovery is crucial to supporting software maintenance and evolution. The clustering problem that could implement architecture recovery is considered as optimizing problem in this paper. Through improving important parameters and core steps of general genetic algorithm, such as initial population, select operator, self-adapting ability of crossover probability and mutation probability, a hybrid genetic clustering algorithm (HGCA) is designed and implemented. An experiment is given to analyze the availability, effectiveness and synthetical performance of the algorithm. The results show that compared to general GA, the HGCA can produce good initial population, better convergence efficiency and convergence precision. Moreover, the value of the MoJo similarity metrics presents the correctness and effectiveness of HGCA recovering software architecture.

Abstract:Any complex system must be controlled by some basic laws, including macroscopic level submicroscopic level and microscopic level laws. How to discover its necessity-laws from these contingency phenomena (observed data) is the most important task of data mining (DM) and KDD, and it is the goal of this paper too. Based on the evolutionary computation and natural fractal, a multi-scale dynamic prediction system is proposed, which models the macro-behavior of the system by ordinary differential equations while models the micro-behavior of the system by natural fractals. The financial data such as the stock market data of Jun抋n stock price and the scientific observed data such as rainfall data of Wuhan in flood season are used as the test data for simulated test of analysis and prediction. The experimental results show that this system fits the data very well, and the simulated prediction is good too, even for modeling the time series with large undulating.

Abstract:Through employing a new error function capturing the nature of multi-instance learning, a neural network based multi-instance regression algorithm is presented in this paper. Experiments on benchmark data sets show that this algorithm deals well with the multi-instance regression problems.

Abstract:A new approach is presented to handle constraints optimization using evolutionary algorithms. It neither uses any penalty function nor makes distinction between feasible solutions and infeasible solutions. The new technique treats constrained optimization as a two-objective optimization. One objective is original objective function, the other is the degree violating the constraints. Individual抯 ranking procedure is based on its Pareto strength which appears first in multi-objective optimization. Pareto strength evaluates an individual抯 fitness dependent on the number of dominated points in the population. By combining Pareto strength ranking procedure with minimal generation gap modal, a new real-coded genetic algorithm is proposed. The new approach is compared against other evolutionary optimization techniques in several benchmark functions. The results obtained show that the new approach is a general, effective and robust method. Its performance outperforms some other techniques. Especially for some complicated optimization problems with inequality and equality constraints, the proposed method provides better numerical accuracy.

Abstract:A range image segmentation algorithm based on Gaussian mixture model of surface normal is proposed. It decreases the times of clustering computing by fully utilizing the physical meaning of Gaussian mixture model of surface normal, and achieves automatic model selection via the posterior probabilities computed from the model parameter estimated by Expectation-Maximization (EM) algorithm. Experimental results on 60 real range images from two kinds of range cameras are compared objectively with some popular segmentation algorithms.

Abstract:Range query is a very important operation to support On-Line Analytical Processing (OLAP) in data warehouses. Although several cube storage structures for range sum queries and dynamic updates have been introduced recently. However, the complexities of both space and time are too higher to realistic. To solve this problem, a hierarchical data cube (HDC) and corresponding algorithms are provided in this paper. Both of the range query and update costs of HDC are O(log^dn), and the overall cost is O((logn)^2d) (under the C_qC_u model) or O(K(logn)^d) (under the C_qn_q+C_un_u model). The analytical and experimental results show that the costs of HDCs range queries, dynamic updates, storage space and the overall performance of HDC are superior to other cubage storage structures.

Abstract:In recent years, the explosively growing amount of data in numerous clustering tasks has attracted considerable interest in boosting the existing clustering algorithms to large datasets. In this paper, the mean approximation approach is discussed to improve a spectrum of partition-oriented density-based algorithms. This approach filters out the data objects in the crowded grids and approximates their influence to the rest by their gravity centers. Strategies on implementation issues as well as the error bound of the mean approximation are presented. Mean approximation leads to less memory usage and simplifies computational complexity with minor lose of the clustering accuracy. Results of exhaustive experiments reveal the promising performance of this approach.

Abstract:Many methods for mapping XML to relations have been proposed without considering the semantics of XML data before. But the semantics is very important to design schemas for storage, optimize queries, and check update anomalies, etc. In the presence of XML FDs which are specified over DTDs, this paper presents a method based on the hybrid inlining to map XML to relations for storage. The constraints which are represented by FDs, as well as the content and the structure, are preserved at the same time. Much storage redundancy can be reduced through using this method. Furthermore, this paper proves that those relations mapped from XML are all in Third Normal Form (3NF).

Abstract:Sensor network, which is made by the convergence of sensor, micro-electro-mechanism system and networks technologies, is a novel technology about acquiring and processing information. In this paper, the architecture of wireless sensor network is briefly introduced. Next, some valuable applications are explained and forecasted. Combining with the existing work, the hot spots including power-aware routing and media access control schemes are discussed and presented in detail. Finally, taking account of application requirements, several future research directions are put forward.

Abstract:An algorithm based on communication algorithm in the Mogent system is presented in this paper. In this algorithm, address registration and multicasting are used by means of address-book so that it is more efficient, and suitable for many kinds of patterns of migration and communication and has a better solution to the troubles with mobile Agent communication.

Abstract:With the rapid growth of network applications, network security has become an important issue. In this paper, the following issues are investigated: the design principles of security protocols, the use of formal methods in the design of security protocols, the characteristics of various approaches to formal analysis, in particular logical analysis. The strand space approach to logical analysis, and the possibility of the strand space model guiding the formal design of security protocols are also explored.

Abstract:In this paper, a Guard Channel (GC) scheme based on dynamic thresholds is proposed. It keeps the ratio of new connection blocking probability to handoff connection blocking probability to a predefined value, which guarantees the priority and fairness among connections. At the same time, it optimizes and improves the performance metrics for handoff: new connection blocking probability and handoff connection blocking probability and system utilization. In addition, the algorithm adopts self-adjusting to be adaptive to varying traffic loads and makes the blocking probability smoother. Lastly the results by analysis and simulation prove the feasibility of this scheme.

Abstract:In this paper, the metrics and principles of testing Web server performance are summarized, and a more rational metric, weighted response time, is proposed. A tool called WebMark is presented for testing Web server performance, which manages asynchronous I/O by event-driving mode and emulates Internet environment by modifying the TCP/IP stack of clients. The testing of the overall performance of Apache proves that developing such a powerful testing tool as WebMark is of great necessity.

Abstract:With the rapid progress of multicast applications and more users participating in it, IP multicast has been an important subject in both research and development. In the meantime, due to the rapid promulgation of mobile devices and progress in wireless technology these years, it is imperative to determine the best way to provide services to mobile equipment. Because IP multicast naturally has the merit of high scalability and efficiency, which is more significant for mobile environment whose resource is much limited, more and more people pay attention to these two techniques, and many efforts are being made to bring them together. In this paper, a comprehensive survey on mobile multicast arithmetics and schemes is given, the advantages and application range of these arithmetics or schemes are discussed respectively, also the shortcomings and problems existed are anatomized in depth, and then a comprehensive comparison is given. In view of poor reliability in mobile environment, a detailed investigation of new problems when implementing reliable multicast to mobile hosts is given, and some related works are introduced. It was hoped that the illustration and discussion presented here would be helpful for developers in selecting an appropriate mobile multicast arithmetic or protocol for their specific needs. In the end, the future trend of combining multicast and mobile is discussed.

Abstract:The issue of reliable communication resulting from Agent moving is studied in the paper. Based on analyzing current solutions to the issue, a mechanism called Active Communication Mechanism (ACM) is proposed, which can provide reliable and efficient communication for mobile Agents. The detailed analysis regarding the reliability and efficiency of ACM is also presented. ACM can reliably and efficiently deliver the sender's message to receiver allowing the free migration of both them and guaranteeing the exactly-once semantics for message delivery. ACM can serve as the basis of high-level communication mechanisms such as Agent communication language.