Applied Mechanics and Materials Vols. 687-691

Abstract: We have designed and implemented a location service framework in the private cloud storage system based on logical disk. Location service is very important in the private cloud storage system and many people are eager to improve the performance of the location service. But the algorithm for location service is fixed so that we can’t change the algorithm even we find a better one. In order to adopt the new algorithm, we need to redesign the location service system. Therefore, we design and implement a location service framework to solve the problem. The location service framework makes development more fast and flexible. The developers and designers can concentrate on algorithm and need not to worry about redesigning the location service system. What’s more, the location service framework can provide the fundamental features of the location service. At last, we prove the effectiveness through simulation testing.

Abstract: To solve the high startup delay and playback delay in P2P streaming system which are caused by peers randomly join or leave the network, we proposed a neighbors acquiring method. This paper presents the details of our neighbors acquiring method. We use tracker server to grasp the information of all nodes in the P2P network. Based on these information, tracker server could pass proper neighbor nodes to the peer node which request neighbors from tracker server through evaluation process. Some conditions are involved in the evaluation process, like node activeness, bandwidth, time nodes existing, buffer-map matching. We estimate our neighbors acquiring method through a P2P streaming system simulation. The simulation results show the P2P streaming system which uses our neighbors acquiring method has lower startup delay and playback delay than the P2P streaming system which uses random neighbors acquiring method.

Abstract: The block cipher New Structure II was designed at eprint 2011, it could be treated as a generalized Feistel structure. By checking the property of the linear permutation, we find a full round differential in New Structure II with probability 1. Since this differential can not distinguish the correct unknown key from the wrong keys, we fail in launching a key-recovery attack on this structure. However, we may guess 1/4 part of the plaintext without calculating the key. And we notice this weakness may cause insecurity in some special environments.

Abstract: The Dual-Channel Network proposed in our previous work can improve the code resource utilization efficiency. Carrier Sense Random Packet (CSRP) CDMA and Partitioned Slot Reservation Packet (PSRP) CDMA were proposed for enhancing the control channel interference resistance of Dual-Channel Networks. In this paper, we investigate CSRP-CDMA and PSRP-CDMA, and derive the performance analytic models by taking into account both intra-cluster MAI and inter-cluster MAI. The models provide the expressions of packet transmission probability and normalized network throughput. Furthermore, the probability of packet successful transmission and the throughput performance are compared between the two protocols. Finally, we analyse the influence on the network performance with spread spectrum gains and network scale. The numerical results show that, normalized network throughput of PSRP-CDMA outperforms CSRP-CDMA by 20% to improving the code resource utilization efficiency of control channel of Dual-Channel Networks.

Abstract: NAT (Network Address Translation) is the process of transforming one IP address in the datagram header into another. In practical applications, NAT is mainly used to realize the function of the access of private network to public network. The method of using a small number of public IP addresses representing a large quantity of private IP addresses will help to slow down the depletion of the IP address space available. As the Network Architecture is becoming more complex, the way, under such context, the NAT technology functions is of great importance. This dissertation will mainly analyze the special treatment of NAT in the following aspects like unified address management ,the priority level of address pool ,NAT mapping in PING Operation ,the treatment of ICMP err packet ,the NAT transformation of fragmented packet as well as the infinitive connection of multi –core products.

Abstract: One single machine learning algorithm presents shortcomings when the data environment changes in the process of application. This article puts forward a heteromorphic ensemble learning model made up of bayes, support vector machine (SVM) and decision tree which classifies P2P traffic by voting principle. The experiment shows that the model can significantly improve the classification accuracy, and has a good stability.

Abstract: Database is the main platform of data management and processing at present, which changes the traditional database management and improve the data management efficiency. However, with the increasing amount of data processing, database faces many problems, data storage capacity expansion, analysis of some data call, etc. In the information age, a mass of data causes the deficit of database management. Therefore, this paper analyzes data management of the database in perspective of capacity expansion and high-efficient call and put forward effective solutions.

Abstract: The progress of computer and network technology provides the technical support for the information construction work in universities, along with the attention of the physique for the students in our country. In this paper, data mining technology combined with the universities sports teaching practice can achieve a sports performance management system based on data mining, and improve the quality of physical education teaching. Through the current situation of universities sports management and data mining can build universities sports management information system, the related value analysis and methods is important to build universities sports management information system based on data mining. The system uses the thought of data mining, which can provide the decision-making basis for sports teaching; In addition, the system structure of three layers makes the system in good adaptability and function expansibility with reduction of the maintenance workload for later.

Abstract: With the rapid development of Internet technology and wide use of computer, digitalization also gets large growth attributes to the surprisingly fast development of science and technology at college. The data sharing has become the bottleneck of digital campus development between the financial system and the use of digital campus data files. In this paper, we make the analysis financial data of the financial system and the digital campus network data sharing, automated processing strategies, and achieve the financial data subscription, automated data extraction. Combining financial system and informatization together with the purpose of realizing a complete sharing of financial information, we have designed the college financial management system, which includes the subsystems: account management, student fees management, compensation management and so on. This paper also makes firstly introduction of the technology adopted in the system and explain the function of each subsystem.

Abstract: With the rapid development of computer technology and network technology, mass data store distributed and management pattern already received accepted extensively. Thus it can be seen malpractice obviously, data storage structure, storage environment are different and other problems such as data handing. The paper go into how improve data storage performance in the distributed environment, analysis the data storage technology at present and data storage performance in the distributed environment, summarize the claim of distributed storage database design, provide the theory in vacation distributed data storage performance standardization.