基于异构节点和时延模型的应用层组播树构建算法研究

发布时间：2018-04-20 01:10

本文选题：应用层组播 + 稳定性因子　；参考：《华中师范大学》2016年硕士论文

【摘要】：当前,移动互联网技术飞速发展和快速推进,推动了移动终端的综合性能快速增加、4G网络商业化部署推进与终端设备功能逐步完善,推动终端用户指数级增加,同时终端用户对移动流媒体需求也快速增长。因此,在移动互联网大发展的背景下为用户提供稳定高效的流媒体服务变的至关重要。大多数移动流媒体应用都需要通过构建传输数据的组播树,复制源节点的数据,进行数据包的发送,为其他节点提供高效的服务。如何构建一个具有稳定的数据传输能力而且时延比较低的组播树显的尤为重要。移动流媒体应用的重点是将移动互联网的处理能力与流媒体的传输和处理技术结合起来,开发出高效的移动应用产品,应用层组播协议作为移动流媒体应用软件的关键技术,与传统IP组播网络相比,不需要底层路由器支持协议,只需要终端节点负责复制与转发数据,使用非常灵活、商业化成本低、方便大规模部署与推广等优点,但是移动客户端具有稳定性差、传输时延大、传输不可靠、网络拓扑结构频繁变化、存储资源受限等缺点,使得应用层组播的使用受到很大约束,因此,如何降低组播协议的传输时延以及构建高稳定性的应用层组播树成为当前应用层组播协议亟待解决的难题。针对上述问题,文章以基于异构节点来构建高稳定性组播树和降低组播树传输时延为目标来完善应用层组播协议,文章的创新性研究表现在下述两个方面：首先,提出一种通过组播节点性能和在线时间运算来进行的应用层组播构建算法HSHN(High Stability based on Heterogeneous Nodes)和备份关键父节点的应用层组播树恢复方法。首先建立模型,构建基于异构节点的高稳定性组播树和提出组播树恢复方式,然后分析模型,论证方法,保证组播树有比较高的稳定性。仿真结果表明,提出的高稳定性HSHN算法能够快速的构建组播树,具有较低的节点加入时延,而且能够快速恢复组播树。其次,定义应用层组播协议的传输时延模型,了解应用层组播传输时延产生的原因以及影响应用层组播时延的关键因素,发现终端节点的处理能力、应用层组播树的是否稳定、节点的度大小以及组播树深度带来的传输路径长短等都影响组播树的传输时延,根据这些特点,文章提出一种利用节点的在线时间与处理能力来决定节点度的算法MDBT (Minimum Delay Balance on Application Layer Multicast Tree),利用时延去构建具有较低时延而且负载均衡的应用层组播树,通过仿真测试,证明MDBT算法构建的应用层组播树能够以较低的时延进行高效的数据传输。
[Abstract]:At present, the rapid development and rapid advancement of mobile Internet technology has promoted the rapid increase in the comprehensive performance of mobile terminals, the rapid development of commercial deployment of 4G networks, the gradual improvement of terminal equipment functions, and the increase of end-user index levels. At the same time, the end user demand for mobile streaming media is also growing rapidly. Therefore, it is very important to provide stable and efficient streaming media services for users in the context of the great development of mobile Internet. Most mobile streaming media applications need to construct multicast tree to transfer data, copy the data of source node, send data packets, and provide efficient service for other nodes. It is very important to construct a multicast tree with stable data transmission capability and low delay. The key point of mobile streaming media application is to combine the processing ability of mobile Internet with the transmission and processing technology of streaming media, and develop an efficient mobile application product. The application layer multicast protocol is the key technology of mobile streaming media application software. Compared with the traditional IP multicast network, there is no need for the underlying router to support the protocol, but only the terminal node is responsible for copying and forwarding data. It is flexible in use, low in commercialization cost, and convenient for large-scale deployment and promotion. However, mobile client has some disadvantages, such as poor stability, long transmission delay, unreliable transmission, frequent changes in network topology, limited storage resources, and so on, which makes the use of application-layer multicast subject to great constraints. How to reduce the transmission delay of multicast protocol and to construct a stable application-layer multicast tree has become a difficult problem to be solved in current application-layer multicast protocols. Aiming at the above problems, this paper aims at constructing a high stability multicast tree based on heterogeneous nodes and reducing the transmission delay of the multicast tree to perfect the application-layer multicast protocol. The innovative research of this paper is as follows: first, This paper proposes an application-layer multicast tree recovery algorithm, HSHN(High Stability based on Heterogeneous nodes, which is based on multicast node performance and online time operation, and backup key parent nodes. Firstly, the high stability multicast tree based on heterogeneous nodes is constructed, and the restoration method of multicast tree is proposed. Then, the model is analyzed and the method is demonstrated to ensure the high stability of multicast tree. Simulation results show that the proposed high stability HSHN algorithm can quickly construct multicast tree with low node delay and can quickly restore multicast tree. Secondly, we define the transport delay model of application layer multicast protocol, understand the cause of application layer multicast transmission delay and the key factors affecting application layer multicast delay, find out the processing power of terminal node and the stability of application layer multicast tree. The degree of the node and the length of the transmission path brought by the depth of the multicast tree affect the transmission delay of the multicast tree. In this paper, a new algorithm, MDBT minimum Delay Balance on Application Layer Multicast tree, is proposed to determine node size by using the on-line time and processing ability of nodes. The application layer multicast tree with lower delay and load balance is constructed by using delay. It is proved that the application layer multicast tree constructed by MDBT algorithm can transmit data efficiently with low delay.
【学位授予单位】：华中师范大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP393.03

【参考文献】