协同化移动流媒体系统中QoS感知的优化调度技术研究

发布时间：2018-04-22 02:39

本文选题：混合网络 + 移动自组网　；参考：《南京大学》2014年硕士论文

【摘要】：近年来,智能手机行业的高速发展见证了移动互联网的蓬勃繁荣。手机作为第一大上网终端,已成为中国网民增长的主要驱动力,并促成移动流媒体服务跃升为移动互联网第五大应用。然而传统基于基站接入的流媒体构建方案因蜂窝网络容量有限、下行速度抖动和资费昂贵等原因限制了移动流媒体的应用普及。为此近年来学术界提出基于蜂窝网络及WiFi网络混成融合架构的移动流媒体系统构建方案,即移动用户一方面通过B2D连接(蜂窝网络接口)与视频服务器连接,一方面通过D2D连接(WiFi)与邻近用户组成移动自组网(MANET)以共享数据。混合网络的P2P移动流媒体系统主要分为两类：第一类为基于单跳广播的P2P移动流媒体系统。系统用户基于P2P协同技术分别下载视频数据的不同部分,然后通过无线网络MAC广播进行节点间数据的分享。系统的优点为数据传输速度快,传输效率高,缺点为用户必须都处于彼此的通信覆盖范围,限制了系统的规模。第二类为基于多跳路由的P2P移动流媒体系统。系统中用户首先尝试通过D2D连接利用Ad-Hoc多跳路由机制从周边节点获取数据,失效时通过B2D连接从远程流媒体服务器获取数据。系统的优点为可以覆盖更多的用户和资源,缺点为数据传输会出现带宽瓶颈,且系统的蜂窝网络负载存在优化空间。在基于多跳路由的P2P移动流媒体系统中数据调度链路的实际拓扑,包括链路长度,转发节点的负载决定了调度的性能,当数据需要经过多个节点进行转发时或转发节点负载很大时,数据调度会出现带宽瓶颈,影响系统的服务质量;其次蜂窝网络负载仍存在优化空间,用户存在重复下载同一数据的现象,且用户的蜂窝数据流量不平衡,影响系统的可扩展性和实际应用。针对上述问题论文对基于旁助节点’协同的流调度优化机制和QoS保障的蜂窝网络负载优化策略开展了研究。本文的主要工作如下：1.本文针对基于智能手机自组网的P2P移动流媒体系统中的数据调度出现的带宽瓶颈问题,提出了基于旁助节点协同的流调度优化机制,通过将系统中的用户和D2D连接建模为加权有向图,将数据调度优化问题规约为最短路径问题,提出了基于Dijkstra算法的最短路径算法,最后通过实验验证了算法随着节点规模的扩大性能稳定优良,具有移动的可扩展性,且算法在不同节点密度,不同旁助节点比例和不同旁助节点缓存上限情况下都能有稳定且优良的性能。2.针对基于多跳路由的P2P移动流媒体系统中蜂窝网络负载优化的问题,在保证QoS的前提下,将此问题建模为求最小下载节点集合的问题,并证明了该问题是一个NP-Hard问题。在此基础上提出了优化蜂窝网络负载和平衡节点蜂窝网络下载数据量的启发式贪心算法。通过实验的验证,算法可以有效的降低系统整体的蜂窝网络流量和平衡用户节点的蜂窝网络下载数据量。
[Abstract]:In recent years, the rapid development of the smartphone industry witnessed the flourishing of mobile Internet. As the largest Internet terminal, mobile phone has become the main driving force for the growth of Internet users in China, and has promoted mobile streaming media services to become the fifth largest mobile Internet application. However, the traditional streaming media construction scheme based on base station access limits the popularity of mobile streaming media applications due to the limited capacity of cellular networks, downlink speed jitter and high tariff. In recent years, the academic community has proposed a mobile streaming media system based on the hybrid architecture of cellular network and WiFi network, that is, mobile users connect to video server through B2D connection (cellular network interface). On the one hand, a mobile ad hoc network (Manet) is formed with neighboring users via D2D connection WiFi. to share data. P2P mobile streaming media systems in hybrid networks are divided into two categories: the first is a P2P mobile streaming media system based on single hop broadcasting. The users of the system download different parts of video data based on P2P collaboration technology, and then share the data between nodes through wireless network MAC broadcast. The advantages of the system are high data transmission speed, high transmission efficiency, and the disadvantage is that the users must be in the communication coverage range of each other, which limits the scale of the system. The second is P2P mobile streaming media system based on multi-hop routing. In the system, users first try to obtain data from peripheral nodes by using Ad-Hoc multi-hop routing mechanism through D2D connection, and obtain data from remote streaming media server by B2D connection when failure occurs. The advantage of the system is that it can cover more users and resources, but the disadvantage is that the bandwidth bottleneck will occur in the data transmission, and there is optimization space for the system's cellular network load. In P2P mobile streaming media system based on multi-hop routing, the actual topology of the data scheduling link, including the link length and the load of the forwarding node, determines the scheduling performance. When the data needs to be forwarded through multiple nodes or when the load of the forwarding node is very heavy, the bandwidth bottleneck will appear in the data scheduling, which will affect the quality of service of the system. Secondly, there is still optimization space for the cellular network load. The phenomenon that users repeatedly download the same data, and the user's cellular data flow is unbalanced, which affects the scalability and practical application of the system. In order to solve the above problems, this paper studies the flow scheduling optimization mechanism based on the side help node 'cooperation and the load optimization strategy of the cellular network supported by QoS. The main work of this paper is as follows: 1. Aiming at the bandwidth bottleneck of data scheduling in P2P mobile streaming media system based on smart phone ad hoc network, this paper proposes a flow scheduling optimization mechanism based on side-help node collaboration. By modeling the user and D2D connection in the system as weighted directed graph, the data scheduling optimization problem is reduced to the shortest path problem, and the shortest path algorithm based on Dijkstra algorithm is proposed. Finally, it is verified by experiments that the algorithm has good stability and scalability with the expansion of node size, and the algorithm has different node densities. Under the condition of different ratio of auxiliary nodes and upper limit of buffer of different side aid nodes, stable and excellent performance. 2. 2. Aiming at the problem of load optimization of cellular network in P2P mobile streaming media system based on multi-hop routing, this problem is modeled as the problem of finding the minimum download node set under the premise of guaranteeing QoS, and it is proved that this problem is a NP-Hard problem. On this basis, a heuristic greedy algorithm is proposed to optimize the load of cellular network and balance the download data of node cellular network. The experimental results show that the algorithm can effectively reduce the overall cellular network traffic and balance the download data of the user nodes in the cellular network.
【学位授予单位】：南京大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN929.5

【相似文献】