P2P流媒体点播系统中基于用户行为特征的缓存策略

发布时间：2018-05-26 04:51

本文选题：P2P + 基于P2P的流媒体点播系统　；参考：《华中师范大学》2015年硕士论文

【摘要】：随着用户对流媒体需求的提高、用户数量的增加,传统的C/S架构模式很容易造成性能瓶颈,影响流媒体技术的进一步发展和广泛应用。传统缓存策略并没有将用户的行为考虑在内,为了更好地支持流媒体点播系统,本文在研究缓存策略时更多地从用户行为特点方面去考虑如何设计缓存策略,所以在新的缓存管理策略中,本文将用户的点播行为特点融入到缓存策略中,使得流媒体点播系统中数据的缓存更加合理,具体研究工作总结如下：(1)介绍了P2P的概念和流媒体的基本概念之后,又介绍了影响视频流畅播放的几个因素。而由于缓存管理策略的不合理,造成了首启动播放延迟,网络抖动等现象。为了减少用户观看视频节目时的等待时间,让系统中的用户节点更加稳定地存在于系统中,需要设计可高效利用缓存空间的缓存策略。(2)由于缓存空间中存储和替换数据块不合理,用户进行频繁地操作会更加影响缓存空间的频繁换进换出,这是因为没有有效地管理缓存空间。由于流媒体点播系统中用户的交互能力大大提高,传统的缓存算法并没有将这一因素考虑在内,因此本文在设计缓存策略时,在之前缓存策略的基础之上,着重考虑用户操作行为的特点。(3)用户在流媒体点播系统中可以进行VCR(快进、快退和暂停等)操作,以及用户可随时加入或退出系统,这些都是研究用户行为特点的关键方面。本文在深入分析这些因素的基础上,提出了数据块的流行度和供求值的概念,基于数据块的流行度和供求值而提出了相应的缓存替换策略。(4)由于现实环境比较复杂,为了验证本文所提出的基于用户行为特征的缓存策略是否合理,本文采用现下比较流行的仿真软件进行验证,实验得出的结论是基于用户行为特征的缓存策略整体上是优于传统策略的。
[Abstract]:With the increase of users' demand for media and the increase of the number of users, the traditional C/S architecture model easily causes performance bottlenecks and affects the further development and wide application of streaming media technology. The traditional caching strategy does not take the user's behavior into consideration. In order to better support streaming media on demand system, this paper studies the caching strategy. In the new cache management strategy, this paper integrates the characteristics of the user's on-demand behavior into the caching strategy and makes the data caching more reasonable in the streaming media on demand system. The specific research work is summarized as follows: (1) the concept of P2P and the streaming media are introduced. After the basic concept, it also introduces several factors that affect the flow of video, and because of the irrational caching management strategy, it causes the first start play delay, network jitter and so on. In order to reduce the waiting time for the user to watch the video program, the user nodes in the system are more stable in the system and need to be designed. Caching strategy for efficient use of cache space. (2) because of the unreasonable storage and replacement of data blocks in the cache space, the frequent operation of the user will affect the frequent change of the cache space. This is because the cache space is not managed effectively. Because the user's interaction ability in the streaming media on demand system is greatly improved, the traditional caching can be used. The algorithm does not take this factor into account, so in this paper, in the design of caching strategy, based on the previous caching strategy, the characteristics of user behavior are taken into account. (3) users can perform VCR (fast forward, fast back and pause) in streaming media on demand system, and users can join or exit the system at any time. These are all Research on the key aspects of user behavior characteristics. On the basis of in-depth analysis of these factors, this paper proposes the concept of data block popularity and supply and demand value. Based on the popularity of the data block and the value of supply and demand, the corresponding cache replacement strategy is proposed. (4) because the real environment is more complex, in order to verify the user behavior proposed in this article. The characteristic caching strategy is reasonable. This paper uses the popular simulation software to verify it. The conclusion is that the caching strategy based on user behavior is better than the traditional strategy.
【学位授予单位】：华中师范大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN919.8

【参考文献】