WDM局域网调度多播业务随机算法的研究

在一个波分复用（WDM）的广播网中,寻找一种最优化消息调度方案比较困难。提出两种调度方案：第一种方案为连续重传一条消息,直到它的所有目标节点都接收到该消息;第二种方案为在两次重传消息之间引入一个随机时延。通过模拟显示,在第二种方案中系统具有更高的吞吐量。     

WiMAX多播广播业务调度算法探讨

本文首先分析了WiMAX中的多播广播业务机制,对MBS网络模型做了简要的介绍。基于该机制,针对不同的场景,提出了两种保证MSQoS的单播多播业务联合调度算法。然后两种算法依据相似的原理将所有的终端划分到不同的调度集合中,针对不同的调度集合分别采用不同的调度策略,以尽量降低终端的能耗。在实际情况中,具体选择哪种算法,取决于MS同时接收多播业务与单播业务数的比例。     

基于等价网络的WDM光网络多播路由与波长分配算法

针对节点功能不同光网络中的路由与波长分配问题,建立了一种具有节点功能区分的波分复用器(WDM)多波长光网络模型,根据节点功能将其分为A,B两类,并提出了波长等价弧、等价网络,以及等价多播树等概念.在此基础上基于节点功能区分的WDM光网络的特点和相应的路由和波长分配策略,提出了一种如何在此类WDM光网络中实现多播连接的快速有效的算法,分析了算法计算复杂度,指出了算法的优缺点.     

有限波长转换能力的OBS网波长分配算法研究

结合国内外最新的研究动态,针对有限波长转换能力的光突发交换网中亟待解决的突发包资源竞争问题,探讨了各种传统的波长分配算法,并通过对其特点的分析和综合比较,提出了一种新的适合有限波长转换能力光突发交换网特点的动态波长分配算法,此算法可以实现更低的全网突发包阻塞率.     

基于混合蚂蚁算法的QoS多播路由算法研究

在研究了QoS多播路由算法的基础上,提出了一种基于混合蚂蚁算法的QoS多播路由算法--QMRA-GAA算法.QMRA-GAA算法利用混合蚂蚁算法构造多播树,真正做到了协议无关,同时QMPA-GAA算法还具有额外负载低、自适应性强、不产生回路等优点.仿真结果表明,和传统的QOS多播路由算法相比,QMRA-GAA算法在加入成功率、自适应性方面都有了很大的提高,建树代价也控制在很好的范围内.     

一种新的应用于光突发交换网络保证流量均衡的波长调度算法

在现有波长调度方法(LAUC_VF)的基础上,结合跳数优先级技术和动态资源分配技术,提出了一种新的可以提供流量均衡的波长调度算法(HP_JET)。通过仿真,证明这种波长调度算法在改善全网阻塞率以及均衡流量方面比原有的波长调度算法有更好的性能。     

一种改进的光突发交换调度算法研究

数据信道调度是OBS核心路由器的关键技术之一,文章对数据信道的几种调度算法进行了分析.在此基础上,介绍了一种改进的波长调度算法.网络仿真结果表明,该算法可以有效改善网络性能.     

加权的基于多播节点的多播路由算法

在许多多播应用中,降低多播树网络费用非常重要.本文提出了加权的基于多播节点的多播路由算法（WDDMC算法）.由于改变了DDMC（Destination-Driven routing for low-cost Multicast ）算法中的指示函数,适当降低了多播节点作为中间节点的优先级,提高非多播节点作为中间节点的优先级,从而使得多播树更接近最小Steiner树.在随机网络上的仿真结果表明,WDDMC算法的多播树网络费用优于DDMC算法.该算法的复杂度与DDMC算法完全相同.     

基于窗口的光突发交换调度算法

针对光突发交换网(OBS)中的资源竞争和QoS支持等问题,提出了一种新型的基于窗口的调度算法。该算法利用边缘节点的电缓存,对调度窗口中的突发数据包进行批量BORA调度,可以有效地降低网络中发生资源竞争的概率;同时调度窗口中的优先级策略还可以进一步增强QoS。网络仿真结果表明,通过选择合适的调度窗口和最大允许时延,该调度算法在0．8的网络负载下使得高优先级突发数据包的阻塞率下降一个量级,信道利用率提高3．5％。     

基于WDM网络的多播动态选路算法研究

提出了一个改进的稀疏分光器配置约束下的改进算法。在稀疏分光器配置约柬下,多播路由不再是图论中“树”的形状,多播路由可能存在逻辑圈。现有的多播路由算法没有考虑这个特点,对此算法作了一下改进,通过仿真证明改进算法提高了网络性能。     

WDM光网络中的组播波长分配算法研究

组播是一种应用广泛的点到多点或多点到多点的通信方式,光层组播以其独特优势引起了人们的关注和重视.在综合分类的基础上,对光网络组播波长分配算法的最新研究进展进行了归纳和总结,并对今后需重点研究的方向进行了展望.     

一种支持扇出分割能力的光组播调度算法研究

组播调度是影响光组播分组交换节点吞吐量的重要因素.针对“一次发送”模型下的光组播调度算法对阻塞的源组播实行全部丢弃的缺点,文章基于节点共享有限波长转换器的光分组交换结构,提出了一种支持扇出分割能力的光组播调度算法,该算法采用随机丢弃阻塞的源组播复制的方法,有效地提高了端口资源的利用率.仿真结果表明,该算法与已有算法相比...     

Study of multicast wavelength arrangement for maximizing network capacity in WDM networks with sparse wavelength converters

The issue of wavelength assignment is one of the most important factors that affect the capacity for the deployment of optical networks. This issue becomes more critical for multicast connections, especially when the network nodes have no wavelength conversion capability. Although the wavelength assignment can be more flexible if each node can perform wavelength conversion, the deployment cost increases accordingly. A compromise is to support a limited portion of conversion nodes in the WDM network. We propose a systematic approach for the wavelength assignment of multicast connections in WDM networks with sparse wavelength conversion nodes. The efficiency of the arrangement of wavelength is measured by its influences on the available capacity of the network and the consumption of wavelengths. By using the proposed approach, the Static Cost Greedy (SCG) algorithm [8] can be easily extended to be applicable in a Sparse Wavelength Conversion Network (SWCN). In addition, instead of SCG, the Minimum-Effect-First (MEF) algorithm is proposed to maximize the network capacity during wavelength assignment. We compare the performance of the proposed MEF methods with the extended SCG scheme through exhaustive simulations. The experimental results indicate that the proposed MEF schemes demonstrate much better performance than the SCG scheme. We also found that the performance is not always improved proportionally to the increment of the wavelength conversion nodes. The improvement reaches saturation when the number of conversion nodes is above 35% of the total number of nodes.

波长转换器的研究

讨论了基于半导体激光放大器的波长转换器，它是光交换块中的关键元件，也是未来通信系统中的关键元件。从理论上分析了各种半导体光放大器转换器的转换原理，并给出了它们的种种结构。     

Dynamic behavior of wavelength converters based on FWM in SOAs

As wavelength converters based on four-wave mixing (FWM) in semiconductor optical amplifiers (SOAs) attract more attention, dynamic effects and wavelength dependent performance become key aspects to be investigated. Such issues are particularly important, as complex configurations are likely to be used to overcome challenges like tunability and polarization dependence. In this paper a numerical model is used to predict the dynamic performance of three FWM configurations and an analytical model is used to derive design rules. First, the wavelength dependent behavior of a wavelength converter is investigated and the requirement for a widely tunable converter is identified. Secondly, a configuration for extinction ratio (ER) improvement is studied and novel design rules are obtained analytically, tested experimentally and explained by the numerical model; experimental results with ER improvement at 10 Gb/s were achieved for the first time. The third configuration studied is a dual-pump arrangement enabling wide tunability. Fixed input/tunable output and tunable input/fixed output configurations are discussed in terms of optical signal-to-noise ratio and tunability. Design rules are extracted and verified for all three configurations that are likely to be deployed: simple wavelength converters, regenerating converters and tunable wavelength converters.     

WDM-PON中一种基于波长选择开关的组播方案

方案使用幅度和相位正交调制的方式来承载单播和组播业务,利用波长选择开关对需要组播业务的ONU(光网络单元)和不需要组播业务的ONU进行分组,进而实现组播业务的传输.整个方案的控制策略完全集中在OLT(光线路终端)一端的WSS中,简化了组播控制的方法,使整个系统简单方便.     

基于时延限制的分布式组播路由选择算法

本文基于最短路径算法提出了主树的概念,并且利用组播树的代价和时延限制的定义给出了一种分布式组播路由选择算法。可以在时延限制的前提下使组播树的代价接近最优。最后,给出了在ＯＰＮＥＴ环境中的仿真结果,并根据结果说明了本文方法的有效性。     

Performance analysis of wavelength converters based on cross-gainmodulation in semiconductor-optical amplifiers

An analytical theory describing all-optical wavelength converters based on cross-gain modulation (XGM) in semiconductor-optical amplifiers is derived. Our theory consists of two parts: a large-signal analysis yielding the transmission function for the signal, and a small-signal analysis in order to describe the transformation of the signal and probe intensity noise. Both the large-signal as well as the small-signal theory reveal similar performance for the co- and the counterpropagating injection scheme for bit rates up to 2.5 Gb/s. This is confirmed by computer simulations. Consequently, the counterpropagating configuration is preferable because the implementation is simpler and conversion to the same wavelength is possible. In order to increase the conversion efficiency it is better to reduce the average signal power than to increase the probe power, which additionally reduces the output power range. However, there is a tradeoff between conversion efficiency and output extinction ratio. According to the small-signal analysis, the relative-intensity noise (RIN) due to the probe and due to the amplified spontaneous emission is negligible. Moreover, the converted signal has a lower RIN than the input signal     

Accurate noise characterization of wavelength converters based on XGM in SOAs

Wavelength conversion (WCR) has recently emerged as an important technique that can be used to manage the spectacular increase in traffic in dense wavelength division multiplexing (DWDM) networks. WCR is extremely useful to solve contention, reduce wavelength blocking and a wide range of WCR methods have been reported. The optimum placement of these devices within a network remains an unexplored area and accurate models to measure the performance of the system are becoming vital. Therefore, this paper presents an original noise characterization of a class of WCRs based on cross-gain modulation (XGM) in semiconductor optical amplifiers (SOAs) when the components from spontaneous emission of the SOAs at the receiver are dominant over thermal and shot noise terms. A new and simple expression for the error probability (P/sub e/) is presented offering considerable additional accuracy in sensitivity assessment compared with the Gaussian approach when ASE noise is dominant. Simulation comparisons are presented in different scenarios for a standard system working at 2.5 Gb/s in a metropolitan area network taking into account the key parameters when such systems are designed.