Advanced Materials Research Vols. 403-408

Abstract: Cloud computing model makes software reuse from the simple code reuse to the service-oriented software reuse .This paper based on the extended Google App Engine cloud computing platform, designed a reusable data reduction tool, which mainly to solve the access problem of heterogeneous data and an un-existed data type. In order to achieve data reuse, a layered data structure definition based on the reuse is provided. Finally, a data-mining system built on top of cloud computing platform is designed and implemented to verify the efficiency and the validity of data reduction module.

Abstract: We present the multi-splitting waveform relaxation (MSWR) methods for solving the initial value problem of linear integral-differential-algebraic equations. Based on the spectral radius of the derived operator by decoupled process, a convergent condition is proposed for the MSWR method. Finally we discussed the convergent condition of discrete-time case of MSWR.

Abstract: In this paper we study the protection mechanisms and security operations of XS128 Flash memory. By introducing the configuration regions in XS128 Flash memory, we elaborate P-FLASH/D-FLASH Protection Mechanisms and protection register FPROT/DFPROT. Besides, we put forward encryption/decryption methods to guarantee the security of MCU.

Abstract: In this paper, for the shortage of network quality of service, multicast support and other deficiencies in VoD systems based on IPv4 network, we built a P2P VoD system architecture based on IPv6 network and data transfer mechanisms were in-depth studied. This paper describes the basic idea of the system and system architecture. According to the characteristics of P2P and multicast, we designed a collaboration data transmission of P2P and IP multicast and did some experiments and analysis on the system. Experimental results show that the system can effectively solve the P2P bandwidth problem and the unreliable of multicast data, and has high efficiency and practical value.

Abstract: This article focuses on FECs which is an error correction coding method and its applications in Internet and wireless communications. First, this article begins with the principles and the types of FECs. Then some typical applications n Internet and Wireless communications for each type of FECs are outlined. The next step is a comparison of the advantages and disadvantages between different types of FECs.

Abstract: The cost estimate plays an important role in cost control and developing new products at the design stage. To improve the accuracy of cost estimate, we extract the feature parameters using the theory of concurrent engineering and factor analysis. Then we propose DCEM that is the model of cost estimate based on factor analysis and BP neural network. The model not only simplifies the input of BP neural network, but also avoids the coupling among the input parameters. The result shows that the model’s performance is stable and it can estimate the cost more accurate at the early product design stage.

Abstract: With the rapid development of information technology, the use of mobile communication tools to infiltrate every aspect of daily life and people is increasingly demanding for the quality of mobile communication network service. As the user base is increasing, mobile communications networks is also increasing pressure. In the traditional artificial methods about communication network fault detection, network managers need to do a lot of work and the process is tedious but not satisfied with the quality test. It is an effective means of fault management using software to manage the network. The mobile network fault detection system base on data mining not only reduces the workload of network managers and the test results more accurate, but also can reduce network management costs. In this paper, we introduce the fault detection strategies for the mobile communication network. By using the data mining algorithm, we can find the network default more easily.

Abstract: Traditional feature weighting algorithm TFIDF doesn’t take some other factors which impact the feature weight into consideration, so this paper discusses the factors in details and proposes a new feature weighting algorithm called NTFIDF combined with these factors and TFIDF. Experiment on the KNN classifier shows that NTFIDF is better than TFIDF in text classification.

Abstract: A surface position location means based on encode recognition is presented; the design of a new kind of absolute position optical mouse using this technology is completed. The position encode algorithm and the code recognition method based on image acquisition and image process are presented also. The structure of pen-type sensor system and light path design of it are discussed. The coordinate conversation algorithm and the detected technology of pen down and pen up are provided. The advantages of the new kind of mouse are that it can write and draw more easily than traditional mouse.

Abstract: Into 2011, the State has strengthened macro-control of real estate, and the policies of restricting purchase have been put forward by the local governments, the real estate industries are facing unprecedented difficulties. The enterprises should adjust their policies of marketing, and actively introduce relationship marketing thinking. To this end, the companies should begin in following eight areas, which change marketing ideas and conception, adjustment corporate governance structure, adjustment structure of human resources, strengthening value management, building long-term external cooperative relation, strengthening service awareness, innovation promotion and sales model, in order to maintain competitive advantage.