【作者】 徐文君；

【导师】 周祖德；

【作者基本信息】 武汉理工大学 ， 通信与信息系统， 2010， 博士

【摘要】 当前,网络技术已经被广泛地应用于多个工业领域,尤其在制造领域中,网络技术的应用和推广取得了令人瞩目的成绩。制造网络可以被认为是通过先进制造模式(如制造网格和云企业等)实现的网络化资源共享管理机制和由Internet实现的计算机通信网络基础架构的集成,因此服务质量(Quality of Service, QoS)保障和优化的观念对于制造网络尤其重要,一方面需要保证用于数据传输和交互的通信网络基础架构的良好运行性能,另一方面还需要对各类网络化的制造资源进行有效管理；以提升相关制造企业的运行效率以及协同和资源共享能力。在制造网络中,如果没有QoS机制的支持,整个网络企业系统的性能都将降低,各类制造资源也无法得到有效利用。针对上述问题,按照“相关理论——关键技术——应用系统”的研究路线,本文对面向制造的网络服务质量相关理论和关键技术进行了深入研究,开展的主要工作如下：(1)讨论了制造网络QoS的概念,建立了制造网络整体QoS保证模型,提出了QoS导向的制造网络服务体系框架。该体系框架集成了网络化资源服务管理和通信网络基础架构两部分的QoS策略,一方面采用蜜蜂优化算法实现对制造资源服务的QoS管理,另一方面在网络基础架构中利用网络QoS策略以满足资源服务管理对于网络传输能力的需求。此外,基于系统自上而下的QoS参数映射机制使整体QoS性能优化的目标得以实现。制造网络中底层网络基础架构是整个系统信息数据传输和交互的前提和支撑,在后续的内容中,主要面向制造中底层网络基础架构的QoS性能优化展开相关关键技术的研究。(2)网络控制系统(Networked:Control System,NCS)是制造网络的典型应用系统之一。NCS中分组传递时延的随机性和动态性对系统控制性能有着较大影响,因此网络分组传递的单向时延估计对于NCS控制性能的优化设计十分重要。分析了当前单向时延估计算法存在的不足,提出了一种适用于NCS、结合在线式测量和端到端估计的混合式单向时延估计算法——HOWDE。该算法针对NCS的运行方式和特点而设计,权对系统中控制启动初始帧的单向时延采用在线式测量方法进行估计,并将其作为端系统运算的参考标量,而对于后续前向和反向通道中的单向时延,则采用端到端的估计方法,基于前期获得的控制启动初始帧参考标量,结合系统两端统计所得的RTT值,通过端系统的运算得出分组传递的单向时延估计值。从而在准确获取NCS中前向和反向通道的单向时延值的同时,不会给系统造成过多的计算开销和网络负担。(3)高速网络是Future Internet的发展方向之一,制造网络中采用高速网络实现底层网络基础架构,可以有效满足日益复杂的制造协同和信息共享对数据传输和交互的高性能需求。在底层通信网络中,用于传输实时数据的UDP协议由于不具有任何控制机制而十分容易引起网络的拥塞,造成共存的TCP数据流性能降低,使整个网络运行的QoS性能下降。针对这个问题,提出了面向高速网络、具有拥塞控制机制的不可靠传输协议——FAST DCCP和EEFAST DCCP。基于DCCP协议并采用FAST控制机制而实现的FAST DCCP协议,通过多Options策略克服了DCCP协议无法在大拥塞窗口下正常运行的缺陷,利用FAST机制实现了对不可靠数据传输的拥塞控制,并且能够区别的对待数据包丢失事件,对丢失的数据包不再重传,以满足数据传递的实时性需求。在FAST DCCP的基础上,提出了加强型端到端协议EEFAST DCCP。该协议基于所测量的单向时延对拥塞窗口大小进行动态控制,从而消除反向流所造成的不良影响。(4)在高速网络基础架构中,现有的高速传输协议在RTT公平性、TCP友好性等网络QoS性能等方面仍存在着缺陷和不足。基于此,提出了采用混合式拥塞控制机制的高速传输协议——HCC TCP。该协议将队列时延作为首要的拥塞度量指标,而将分组丢包作为次要的拥塞度量指标。协议中基于时延的控制机制通过对前向单向时延的统计测量,准确的估计出前向路径中的拥塞状态,从而根据时延信息动态调整拥塞窗口的大小,使其稳定在可以完全利用可用带宽资源的窗口值上。当基于时延的拥塞控制机制由于设置参数的制约而无法在网络中有效运行时,协议则采用基于分组丢包的控制机制对窗口大小进行调整,使窗口的增长呈线性到曲线的变化,从而将窗口能够尽可能长时间的维持在分组丢包事件窗口值附近,降低分组丢包的发生概率,取得良好的带宽利用率。(5)针对制造产业中中小型企业(Small and Medium Size Enterprises, SMEs)所面临的挑战,基于Future Internet提出了能够使SMEs运行于未来企业系统的创新网络环境,同时对该环境的体系结构和相关方法论进行了描述和分析,从而使SMEs能够在快速变化和充满竞争的市场环境中生存并取得成功。最后,以欧盟第七框架计划ICT和NMP中的相关项目为应用实例对本领域的研究进展进行了分析,指出了所提出的创新网络环境较现有研究成果所具有的先进性,以期能够帮助制造企业成功应对未来市场的挑战。更多还原

【Abstract】 Nowadays, networking has been adopted in a variety of industrial fields. In particular, in manufacturing industry, the move towards networking has become a notable trend. Manufacturing networks can be regarded as integrating communication networks and networked resource service management using novel manufacturing models, such as Manufacturing Grid or Cloud Enterprise. Therefore, resolving Quality of Service (QoS) issues is important for manufacturing networks, not only to guarantee the performance of communication networks, but also to achieve effective management of various manufacturing resources. Without such effective QoS mechanisms, system performance will degrade and the manufacturing resources may become unusable. Aiming at the aforementioned problems, the thesis focuses on the relative theories and key technologies of QoS assurance for networks in manufacturing. The main contributions of the thesis are:(1) The concept of QoS for manufacturing networks is discussed, and the related QoS model is established. In order to provide overall performance assurance for manufacturing networks, a service framework integrating the QoS mechanisms of the networked resource service management function and the communication networks is proposed. The novel framework maps an application to resource services and then to communication networks, adopts an intelligent optimisation algorithm for QoS management of resource services,and provides QoS schemes for data transfer across communication networks.The communication network infrastructure in manufacturing networks is responsible for the data transfer and exchange in the system, and therefore it is the basis supporting the operation the networked resource service management. In the following content, the thesis mainly focuses on the network QoS issues in manufacturing networks, and to develop the key technologies in order to provide QoS assurance for the network infrastructure.(2).One-way delay estimation in the forward and feedback channels is crucial for a networked control system (NCS) due to its important role in the reliability and availability design. The problems existing in current one-way delay estimation schemes are investigated. Then, a novel hybrid one-way delay estimation scheme, utilizing the hybrid technique that contains an online monitoring mechanism and an end-to-end estimation method, is proposed to overcome the effects of network asymmetry and delay dynamics in the NCS. In the proposed scheme, only the delay value of the initial packet starting the control loop of the NCS is estimated by the online monitoring mechanism, all the other estimations are performed by the host using the end-to-end method, so as to achieve high accuracy in tracking delay values as well as requiring lower computation complexity of the system.(3) In the high-speed network infrastructure of manufacturing networks, the real-time traffic of multimedia applications can congest the network and result in unfairness and throughput degradation of Transmission Control Protocol (TCP) traffic, in terms of there is no control mechanism in User Datagram Protocol (UDP). In order to improve the QoS performance of the network infrastructure, a new unreliable transport protocol, FAST DCCP, is presented for the real-time traffic in high-speed networks. FAST DCCP is based on the DCCP protocol and adopts the FAST scheme to realize congestion control. Some modifications have been made to the mechanisms inherited from DCCP so as to let the proposed protocol can efficiently operate under a large size window. In addition, an enhanced protocol, EEFAST DCCP, using the measurements of one-way delay to dynamically adjust the window size is also proposed to improve the throughput of FAST DCCP with the effect of reverse traffic.(4) Although several high-speed TCP variants perform successfully to improve the bandwidth utilization of high-speed network, they still have the weakness on the performance such as RTT-fairness, TCP-friendliness, etc. A hybrid congestion control algorithm, HCC TCP, using the synergy of delay-based-and loss-based approach is proposed for the adaptation to high speed and long distance network environments. The algorithm uses queuing delay as the primary congestion indicator, and adjusts the window to stabilize around the size which can achieve the full utilization of available bandwidth. On the other hand, it uses packet loss as the second congestion indicator, and a loss-based congestion control strategy is utilized to maintain high bandwidth utilization in the cases that the delay-based strategy performs inefficiently in the networks. The two approaches in the algorithm are dynamically transferred into each other according to the network status.(5) Facing the the challenges for future SMEs in manufacturing, the thesis presents an innovative network environment, together with its architecture and methodologies, to support the development of Future Internet based Enterprise Systems for the SMEs, in order to achieve the goals of high operation efficiency and sustainable value creation. A list of related EC FP7 projects focusing on the networked enterprise domain and their characterization are introduced, analyzed, and compared. Finally some research issues and future works are given.更多还原