Design And Implementation Of Resource Sharing Platform Based On P2P

In past, network services or resource sharing was completed in the form of C/S(Client/Server). All the resources are stored on the servers of networks. If one client needs to download resources, it is necessary to link to the servers, send request of sharing resources, and wait for the server’s resource transmission. However, under this model, a lot of work is finished on the servers, resulting in servers’ overload. Moreover, the transmission efficiency is heavily dependent on bandwidth. When the users’ requests become more, the download speed becomes slower. More importantly, once the server stops running, it is impossible to share resources. Aiming at the problems of resource and service sharing in C/S mode, P2P(Peer Peer) technology emerges as the times require. The technology completely changed the way of sharing resources by the C/S mode. Previously, all resources are stored on servers. In the P2 P mode, resources can also be stored in the network terminal, and even only in the terminal. Therefore, the pressure on servers is greatly reduced. When a terminal sends request of obtaining resource, a plurality of terminals can respond to the request simultaneously, and the resource is transmitted to the terminal. When the resource acquisition is completed, the terminal can also be used as a temporary "server" to respond to the needs of other terminals. In fact, P2 P technology breaks the boundaries between the client and the server. The client can also become a server at a certain time, which greatly improves the efficiency of resource sharing.Hence, the author developed a set of P2 P based resource sharing platform, hoping that this system can bring convenience to users. The system is conducive to the rapid access to network resources. At the sa me time, personalized recommendation according to the characteristics of users is embedded into the system.In this paper, the P2 P based resource sharing platform developed in this paper mainly consists of three main modules, which are the front-end of the system, the background management module for common users, and the background management module for administrator users. The basic function of the front end of the system is information display. The basic functions of the background management for general users include: user login, registration and exit, seed file management, topic group management, and the message management. Background management module for administrator users includes: seed management, user management and system management. The entities in the system include administrators, users, seeds, seed types, message, groups, download records, group topics, group discussions, website information, recommended lists and hyperlinks. Except the administrator, website information and hyperlinks, other entities are associated with each other.Considering that the file sharing platform contains a large number of files. How to help users find their own interested files in such a large database is one of the focuses in this platform. Recommended system can solve this problem well. The aim of the system is to provide users with personalized recommendation service. The basic principle of the recommendation is to analyze the attributes of the users, the characteristics of the goods, and the relationship between users and objects. Based on the relationship between people and things, the similar relationship between human and human, the relationship between the new user and objects can be derived. In this paper, a new algorithm based on clustering is proposed. The algorithm consists of two parts: user clustering and file recommendation. By observation, it is found that the local density of the instance around the center of the class is lower than that of the center of the class. And the distance from the instance w ith higher local density is far. The local density of each sample and the minimum distance between each instance and the one with higher local density can be calculated. The instance with two higher indicators is the cluster center; if the former indicator is low and the latter is higher, it is a noise point. Other points are normal. The advantages of this algorithm can be summarized as:(1) it is not necessary to determine the number of clusters in advance;(2) there is no need to initialize cluster centers;(3) it does not need iteration, which greatly reduces the time cost. File recommendation strategy uses the K nearest neighbor(KNN) principle. By finding the K the most similar users, video files with higher scores can be found. A list of recommended fi les is given where elements are ordered in accordance with the averaged scores. In addition, this paper also designs a new performance evaluation index of the recommendation algorithm.The structure of this paper is as follows:In the first chapter, the re search background and significance of the resource sharing platform based on P2 P are clarified, the shortcomings of the traditional resource sharing platform are summarized, and the importance of resource sharing platform based on P2 P is fully explained.In the second chapter, the P2 P technology is introduced in detail, including the basic principle of P2 P technology, file location technology, node selection, fault tolerance mechanism, security mechanism and so on.The third chapter describes the requirements of the system, including the project objectives, the overall description, non- functional requirements, etc..The fourth chapter is the overall design of the system, including the analysis of the overall system design, database design, hardware and software environment, etc.In the fifth chapter, a new algorithm based on clustering is designed.The sixth chapter shows the design and implementation of the system, and describes the implementation process of the system in detail.In the seventh chapter, the resource sharing platform based on P2 P is summarized, and the further development direction of the system is given.