自组织网络中调度算法的研究

Scheduling Algorithms for Self-organized Networks

作者: 专业:信息与通信工程 导师:田臣 年度:2016 学位:硕士  院校: 华中科技大学

Keywords

Self-Organized networks, Shadow Stream, Distributed scheduling, Cooperative communication

        近来年随着硬件设备的发展和网络规模的迅猛增长,自组织网络(Self-organized Networks)逐渐渗入社会发展的方方面面。从日常生活中信息的传播,如视频直播、资源下载,到军事场景下即时信息的传播及关键信息的反馈,自组织网络无不发挥着重要的作用。同时,在不同场景下,如何进一步提高自组织网络的性能,满足用户的服务质量(Quality of Service,QoS)也成为越来越多研究者关注的问题。本文针对两个运用在不同场景下的自组织网络中的调度算法进行了研究和优化,希望能进一步提高了特定运用场景下自组织网络的性能,更好地为客户提供服务。随着P2P技术在流媒体在线直播平台上的应用,流媒体平台逐渐成为人们获取信息的重要手段。同时,为了克服传统的软件测试方法带来的测试结果不准确的问题,保证客户的服务质量,嵌入了在线测试功能的流媒体直播平台影子系统应运而生。其中,影子系统中的一个重要组成部分是中心控制点,它主要负责给挑选出来进行测试的客户分配一个虚拟的到达时间和停留时间,以满足开发者指定的测试模式,得到可靠的测试结果。然而,在影子系统最初的设计中仅仅考虑利用网络中小部分相对稳定的客户进入测试阶段,这一设计在很大程度上影响了可以参与测试的客户数目从而限制了测试规模。因此,我们提出了一个新颖的分布式生命周期控制算法,充分利用具有不同剩余停留时间的客户克服网络中稳定客户数目的限制。算法的核心在于利用网络中真实客户实际行为模式和测试过程中预期行为模式的匹配。仿真结果表明在满足预期测试模式的前提下,采用该算法能将测试的规模增大近一倍。此外,由于移动自组织网络自形成和自恢复的特性能很好地适应复杂的军事环境,现代军事通信技术正逐渐将移动自组织网络技术应用到边缘地带的军事战略中。同时,由于军事场景下的应用与一般传统通信的应用大相径庭,移动自组织网络的发展前景也面临着无数的挑战。Barrage Relay Networks(BRNs)作为一种新的军用无线自组织网络能够从根本上满足战术军用网络通信的需求。同时,基于协同通信这一技术BRNs能够在单点广播场景下即时有效地传播数据包。然而,BRNs中的广播机制并不适用于多点同时广播的情景。一种比较简单的做法是采用单点广播的机制让每个源节点依次发送数据包,然而这种方法并不能满足军用场景下对数据时延的要求。因此,我们在BRN的多点广播场景下提出了一个基于网络拓扑结构的中心时隙调度算法。仿真结果表明该算法并不受网络大小的影响,而且与传统的广播机制相比能够将网络的吞吐量提高近23%。
    With the rapid development of hardware devices and increasingly scale of network, self-organized network is widely applied in our daily life and plays a significant role nowadays, ranging from the transmission of daily information, such as live streaming, resources download, to the edge communication in modern military fields. Thus more and more research tend to focus on ways to improve the performance of ad hoc network and guarantee users’ quality of service in different application scenarios. Our work aims at developing efficient scheduling algorithms in two specific scenarios in self-organized network to further ameliorate its performance and provide users with better services.As peer-to-peer technology being integrated in live streaming platforms, streaming service becomes a major Internet application that has been used to carry both daily and major events. However, as live streaming networks continue to grow in scale and complexity, they are becoming increasing complicated to evaluate. Existing evaluation methods including testbed testing, simulation and theoretical modeling, lack either scale or realism, which have greatly affected the accuracy of testing result. Thus, Chen Tian has designed Shadow Stream, a novel Internet live streaming system that integrates performance evaluation as an intrinsic capability. An essential component in Shadow Stream is distributed lifecycle control mechanism, which assigns each client a virtual arrival/lifetime to create a particular scenario to evaluate the performance of streaming system. The original design focuses on utilizing stable streaming viewers in physical world to guarantee the accuracy of Shadow Stream, which, on the other hand, significantly limits the scale of the experiment. Our research develops a novel distributed client lifecycle control to get rid of restrictions caused by the limited number of stable viewers in live-testing streaming networks. The core idea of our research is to match the desired experimental scenario with real viewers’ behavior in physical world. The result demonstrates that with our methodology, the scale of experiments can be doubled.In addition, modern tactical edge communication is turning to mobile ad hoc networks(MANETs) since the self-forming and self-healing nature of MANETs is advantageous in such a complicated military environment. However, MANET faces a myriad of challenges in tactical filed compared with traditional scenarios. Barrage Relay Network(BRN) is a special MANET designed from the ground up for the demands of tactical edge communications, which defines a rapid, robust and scalable broadcast mechanism based on an autonomous cooperative communication scheme. While, there is no research about time slot scheduling in a multiple source broadcast scenario yet. Na?ve approach is to let each source node broadcast a package one by one, which is not optimal in terms of the strict requirements of latency. In this paper, we put forward a centralized topology dependent scheduling algorithm designed for multi-source transmission scenarios in BRNs. Extensive evaluations demonstrate that the performance is not sensitive to the size of network and the throughput of network has been improved nearly 23% when compared with the na?ve approach.
        

自组织网络中调度算法的研究

摘要4-6
ABSTRACT6-7
1 绪论10-18
    1.1 研究背景10-13
    1.2 研究意义13-15
    1.3 论文研究内容15-16
    1.4 论文章节安排16-18
2 P2P自组织流媒体网络中的测试客户调度问题18-25
    2.1 P2P与流媒体技术18-19
    2.2 影子系统基本设计19-20
    2.3 分布式接入控制机制20-22
    2.4 基本思想22-25
3 在线流媒体测试平台中分布式算法设计25-37
    3.1 触发条件25-27
    3.2 到达时间分布式控制算法27-28
    3.3 生命周期分布式控制算法28-31
    3.4 稳定用户的“再进入”31-32
    3.5 仿真实验与结果32-37
4 BRNS中多源广播场景下的时隙调度算法37-51
    4.1 军用自组织网络中的时隙调度问题37-41
    4.2 节点数据结构41-42
    4.3 节点状态的定义42-44
    4.4 多源广播场景下的时隙调度算法44-47
    4.5 仿真实验和结果47-51
5 总结与展望51-53
    5.1 论文工作总结51-52
    5.2 未来研究方向52-53
致谢53-55
参考文献55-61
攻读硕士期间的学术成果61
        下载全文需58


本文地址:

上一篇:基于图像标注的在线协作系统的设计与实现
下一篇:基于SDN的改进延迟调度算法

分享到: 分享自组织网络中调度算法的研究到腾讯微博           收藏
评论排行