一种基于策略函数的应用层组播路由算法

由于IP组播存在可扩展性差、难以管理等方面的缺陷,研究人员提出了应用层组播。实时传输是应用层组播技术的一个主要应用领域,对网络延迟有严格的限制。文中着重研究构建最小延迟应用层组播树的算法,提出一种基于策略函数构造应用层最小直径组播树的启发式算法BCT-H。该算法采用策略函数迭代的选择使生成树直径最短的路径,从而有效地减少了网络中的转发时延和同一条链路的重复分组数量。模拟实验表明该算法能够有效地降低链路强度,减少组播树的时延。     

一种基于策略函数的应用层组播路由算法

由于IP组播存在可扩展性差、难以管理等方面的缺陷,研究人员提出了应用层组播.实时传输是应用层组播技术的一个主要应用领域,对网络延迟有严格的限制.文中着重研究构建最小延迟应用层组播树的算法,提出一种基于策略函数构造应用层最小直径组播树的启发式算法BCT-H.该算法采用策略函数迭代的选择使生成树直径最短的路径,从而有效地减少了网络中的转发时延和同一条链路的重复分组数量.模拟实验表明该算法能够有效地降低链路强度,减少组播树的时延.     

PPNCast：一种基于对等交换的应用层组播方法*

针对高码率和实时性要求不高的大规模媒体分发应用,提出一种新的应用层组播PPNCast。PPNCast基于对等交换的思想对应用层组播系统进行了优化,可以支持多种片的分发模型。通过仿真实验与Yoid、NICE应用层组播协议对比,结果显示PPNCast可以明显改善系统的链路压力均衡度和节点公平度。     

应用层组播共享拥塞链路消除算法

应用层组播网络由于是构造在基础网络之上的覆盖网络，因此在应用层组播路由中产生了共享拥塞链路问题。根据共享拥塞链路的不同特点，提出了一种新的基于组播会话流而改变组播树拓扑结构的应用层组播树生成算法，消除组播树中的共享拥塞链路。仿真实验表明，与基于延迟或带宽的同类路由算法相比，此算法在综合考虑组播树的带宽利用率和延迟条件下具有更好的性能。     

大规模应用层组播网络体系结构研究

阐述了IP组播发展所面临的问题、应用层组播发展现状和ScatterCast应用层组播网络，在对ScatterCast扩展的基础上，提出了一种新的支持大规模应用模式的应用层组播网络体系结构，以解决应用层组播通信的可扩展性及与IP组播的兼容性等问题。介绍了应用层组播网络的性能评价参数。     

基于节点性能估算的应用层组播算法

提出一种基于节点性能估算的应用层组播树构建和维护算法(PPE)。综合考虑节点的性能及其在组播树中的位置,尽可能使高性能的节点接近源节点,使低性能的节点接近叶子节点,以此来提高组播树的效率,降低端到端的延迟,减小链路压力和控制开销。仿真结果表明,PPE在大规模节点环境中能有效改善组播树的性能。     

基于二维网状结构的应用层组播协议研究

为了改善网状优先的应用层组播协议可扩展性差等缺点,提出一种新的应用组播协议,协议采用特殊的二维层次结构。在低层,将距离近的结点划分为同一个域,域内成员组织成车轮状,域成员只和邻居结点交换信息,减少了维护通信量。在高层,将距离较远的域首组织成网状结构,上下层通过不同的路由算法并行地传输数据,使得数据可以尽快到达远处的结点,提高了传输效率。仿真结果表明,该组播协议具有较小的控制开销、链路压力和端到端延迟。     

基于拓扑结构和节点性能的应用层组播算法

针对当前应用层组播节点具有异构性和存在传输延迟等问题,提出一种基于拓扑结构和节点性能的应用层组播算法。根据拓扑感知的方法对网络进行分层分簇,在构建组播树时考虑节点的异构性,使高性能的节点接近源节点,低性能的节点接近叶子节点。实验结果表明,该算法能提高组播树的效率,降低端到端的延迟,减小链路压力和控制开销,在大规模节点环境中有效改善组播树的性能。     

异构网络环境下的高稳定性应用层组播方案

针对异构网络环境中由于节点转发能力异构而引起的应用层组播稳定性不足的问题,对异构overlay拓扑建模进行了研究,分析出组播稳定所需满足的约束条件,提出了一种基于gossip协作机制的应用层组播方案。在所提机制能够均衡节点间负载,缓解能力弱的节点负载过重的问题;如果组播过程中有成员节点动态离开或转发能力不足,其他节点相互协作共同分担其负载以维持组播的稳定性。随后的链路强度分析表明,与传统的应用层组播方案相比,本方案能够大幅降低拓扑的平均链路强度,提高组播系统的稳定性。最后通过OMNeT 仿真平台对本方案进行仿真验证,实验证明在动态的异构overlay拓扑下,本方案能够显著提高数据分发成功率,保证组播的连续性和稳定性。     

基于节点异构性的混合组播网络

为了扩大导航差分信息的应用领域,并且在城市或室内可以接收修正信息,针对导航增强信息需在具有实时性、可扩展性和异构性的网络上播发的特点,结合应用层组播和i3组播,采用IP路由信息表构建了基于节点异构性的分层混合组播系统.分析了网络模型和通信协议,并基于OPNET进行了性能仿真.仿真结果表明,HMMA实时性好、可扩展性强,可以满足应用的需求.     

<Emphasis Type="Bold">Ppssm:push/pull smooth video streaming multicast protocol design and implementation for an overlay network</Emphasis>

IP multicast is one of the best techniques for video streaming on the Internet. It faces issues with respect to address allocation, routing, authorization, group management, security, and scalability. By default, local Internet Service Providers did not enable IP multicast services, because of the cost incurred in using multicast-enabled routers. To solve these issues some of the IP layer functionalities have been shifted to the Application Layer, thus leading to Application Layer Multicast (ALM) protocols. However, ALM protocols face issues related to synchronous data delivery, scalability, link stress, link stretch and node failures. Some of the existing protocols are CoolStreaming, and mTreebone. A novel ALM protocol based Push/Pull Smooth video Streaming Multicast (PPSSM) protocol is proposed in this paper, to increase the throughput and reduce the packet loss rate. The PPSSM protocol involves three stages, such as tree-mesh construction, dynamic buffer management and network coding techniques. In the tree-mesh construction, a tree consists of stable nodes and a mesh consists of unstable nodes. The proposed PPSSM optimizes the stable nodes in the tree, which minimizes or eliminates the pull operations from the unstable mesh overlay nodes, by exploring the potential of the stable nodes. Dynamic buffer management is achieved by setting the optimal buffer threshold value, using the optimization of the sensitivity parameters, such as packet loss and packet workload/delay by the Infinitesimal Perturbation Analysis and Stochastic Approximation algorithms. In addition to the tree-mesh construction and buffer management, the introduction of the network coding technique will enhance the throughput and minimize the packet loss and delay. Finally, the performance of the proposed PPSSM protocol is compared with those of CoolStreaming, and mTreebone, and it shows improvement in respect of throughput, packet loss, and average decoding time.     

基于多组密钥树的应用层组播组密钥管理方案

针对现有应用层组播组密钥管理方案可扩展性差等问题,设计一种适用于大规模应用层组播的基于多组密钥树的组密钥管理方案MGKT。该方案对组成员分簇管理,将组密钥的更新限制在簇的范围以内,从而降低了组密钥管理的通信开销。分析和仿真实验表明,在不降低安全性的情况下,MGKT具有较好的可扩展性。     

ALMware: A middleware for application layer multicast protocols

This paper presents a new middleware called ALMware for supporting development and performance evaluation of ALM (application layer multicast) protocols on real environments. To facilitate the development of ALM protocols, ALMware provides APIs which allow protocol developers to re-use typical operations used in existing ALM algorithms. Some additional APIs for transmitting video and audio between peers also mitigate the cost of writing application codes. Performance evaluation on real environments is also supported by ALMware. To prove the usefulness of ALMware, we have implemented well-known but different types of ALM protocols, ALMI, NARADA, NICE and OMNI, and conducted various kinds of performance evaluation experiments on PlanetLab, which clearly demonstrate the usability of ALMware. We believe that this is the first effort made to develop this kind of ALM supporting middleware.     

一种自组织的应用层组播层次密钥管理方案

针对现有应用层组播密钥管理方案可扩展性差、时延较大等问题,设计一种适用于大规模的自组织的层次密钥管理方案(SOKM).该方案将密钥服务器(KS)的任务分散给各个簇的特殊节点来完成,将组密钥的更新和管理限制在簇范围内.分析和仿真实验表明,与分层分簇方案MHKM相比,SOKM虽然在一定程度上增加了计算量,却有效降低了KS受攻击的可能性和组密钥分发时延,使整个组播组的安全性和密钥分发效率得到了提高,因此具有较好的可扩展性.     

模拟训练中可靠组播技术研究与实现

该文针对舰艇模拟训练系统的需要,提出一种可扩展的可靠应用层组播模型,它采用心跳机制与序列号相结合的方法进行差错检测和基于NAK否定信息的ARQ机制进行数据差错恢复,同时设计了相应的机制来保障组播数据的可靠传输。该模型弥补了普通应用层组播协议在可靠性传输上的不足,使通信网络具有良好的可扩缩性、可选择和可靠性功能。实验证明：基于心跳机制的差错控制模型可以很好克服传统方法可靠组播通信的＂反馈风暴＂和NAK算法中的缓冲区无法控制的问题,与基于TCP的单播可靠方式相比,该模型不需要中心节点服务器,节省了网络资源,网络带宽占有率相比也减少约两倍。     

分层的应用层组播协议研究

在P2P视频直播网络中,用户频繁地加入或离开组播组会造成数据传输中断。该文提出一种新的基于分层分簇思想的应用层组播协议,基于视频直播中的用户行为分析,通过分层获得一个有效层来减少组播树中的节点失效次数。仿真实验表明,该协议能够有效提高稳定性,减少平均组播时延并具有可扩展性。     

Parallel overlays for high data-rate multicast data transfer

Recently a number of application-layer multicast (ALM) protocols have been proposed as a promising alternative to deploying multicast services in the unicast-only Internet. Current ALM protocols work very well for low data-rate applications but can suffer from link-level load imbalance, and consequently link congestions, when applied to high data-rate applications. This work addresses this problem by extending the well-known NICE protocol to use multiple parallel overlays in the same ALM session to spread the data traffic across more available network links, and thus leading to significantly improved performance. Extensive simulation results show that the proposed protocol can support three times the data-rate compared to NICE and yet can reduce the end-to-end data delivery delay by more than 50%.     

On Maximizing Tree Bandwidth for Topology-Aware Peer-to-Peer Streaming

In recent years, there has been an increasing interest in peer-to-peer (P2P) multimedia streaming. In this paper, we consider constructing a high-bandwidth overlay tree for streaming services. We observe that underlay information such as link connectivity and link bandwidth is important in tree construction, because two seemingly disjoint overlay paths may share common links on the underlay. We hence study how to construct a high-bandwidth overlay tree given the underlay topology. We formulate the problem as building a Maximum Bandwidth Multicast Tree (MBMT) or a Minimum Stress Multicast Tree (MSMT), depending on whether link bandwidth is available or not. We prove that both problems are NP-hard and are not ap-proximable within a factor of (2/3 + epsiv), for any epsiv > 0, unless P = NP. We then present approximation algorithms to address them and analyze the algorithm performance. Furthermore, we discuss some practical issues (e.g., group dynamics, resilience and scalability) in system implementation. We evaluate our algorithms on Internet-like topologies. The results show that our algorithms can achieve high tree bandwidth and low link stress with low penalty in end-to-end delay. Measurement study based on Plan-etLab further confirms this. Our study shows that the knowledge of underlay is important for constructing efficient overlay trees.     

应用层组播研究进展

组播技术是一种针对多点传输和多方协作应用的组通信模型,有高效的数据传输效率,是下一代Internet应用的重要支撑技术。早期的组播技术研究试图在IP层提供组播通信功能,但IP组播的实施涉及到对现有网络基础设施的调整,因此,大规模应用受到限制。近两年来,随着Peer-to-Peer(P2P)研究的兴起,基于应用层的组播技术也逐渐受到广泛关注。应用层组播协议将组成员节点自组织成覆盖网络,在主机节点实现组播功能,为数据多点并发传输提供服务。将组播功能从路由器迁移到主机上能有效解决许多与IP组播有关的问题,但同时也带来了一些新的挑战。本文分析了目前应用层组播研究的主要内容及技术特点,描述了协议设计所涉及的关键技术及面临的主要挑战,总结了现有工作及相关进展。