CLEP——一种新的全光网络光路建立协议

本文提出了一种波长交换光网络中混合模式的光路建立协议CLEP,该协议充分利用现有基于正向预留的CR-LDP协议和基于反向预留的RSVP-TE协议各自的特点,通过使用混合模式的光路建立机制,使得信令协议在控制消息负载、实现复杂度、连接阻塞概率和连接建立时间等方面具有更好的性能.     

Fast Routing Algorithms for Lightpath Establishment in Wavelength-Routed Optical Networks

In this paper, we study shortest-path routing in wavelength-routed optical networks with an objective to optimize the average-case running time for path computation. Four fast routing algorithms are proposed for dynamically computing the shortest lightpaths or semilightpaths in a network with or without wavelength converters. To reduce the average-case running time for path computation, sequential search, backward routing, and informed search are used in the algorithm design. Simulation results show that the proposed algorithms can significantly reduce the average-case computational overhead for path computation as compared with existing algorithms.      

Virtual Fiber Configuration for Dynamic Lightpath Establishment in Large-scaled Optical Networks

Recently, progress has been made in the Generalized Multi-Protocol Label Switching (GMPLS) and Automatic Switched Optical Networks (ASON) standardizations. These technologies realize construction of large-scaled optical networks, interconnections among single-domain Wavelength Division Multiplexing (WDM) networks, and direct communication over multi-domain WDM networks. Meanwhile, it is known that the topology of the Internet exhibits the power-law attribute. Since the topology of the Internet, which is constructed by interconnecting ASs, exhibits the power-law, there is a possibility that large-scale WDM networks, which are constructed by interconnecting WDM networks, will also exhibit the power-law attribute. One of the structural properties of a topology that adheres to the power-law is that most nodes have just a few links, although some have a tremendous number of them. Another property is that the average distance between nodes is smaller than in a mesh-like network. A natural question is how such a structural property performs in WDM networks. In this paper, we first investigate the property of the power-law attribute of physical topologies for WDM networks. We compare the performance of WDM networks with mesh-like and power-law topologies, and show that links connected to high-degree nodes are bottlenecks in power-law topologies. To relax this, we introduce a concept of virtual fiber, which consists of two or more fibers, and propose its configuration method to utilize wavelength resources more effectively. We compare performances of power-law networks with and without our method by computer simulations. The results show that our method reduces the blocking probabilities by more than one order of magnitude.     

Scheduling Network and Computing Resources for Sliding Demands in Optical Grids

In this paper, we consider the problem of scheduling lightpaths and computing resources for sliding grid demands in WDM networks. Each sliding grid demand is represented by a tuple (v,R,c,p,q,l) , where v is the client node, R is the resource-group which includes a group of predefined resource nodes, c is the required amount of computing resources, [p,q] is the time window and l is the demand duration. With each demand, the scheduling algorithm is required to decide the start time t (p les t les q - l), reserve an amount of c computing resources at a resource node v ^' isin R and provision a primary lightpath as well as a backup lightpath from v ^' to v . The reserved computing resources and lightpaths are used during [t,t + l]. Unlike the sequential approach wherein the start time, the network resources (lightpaths) and the computing resources are considered one after another, in our work we use the joint scheduling approach wherein the resources and the start time are examined jointly. We consider sliding demands with static and dynamic arrival patterns. We develop an integer linear programming (ILP) formulation to obtain optimal results. For the reason of scalability, we propose heuristic algorithms based on joint resource scheduling and study their effectiveness through simulation experiments.     

ASON中一种基于光波长路由器的快速光路建立方式

通过对目前ASON网络中三种光路建立方式不足之处的分析,提出了一种基于光波长路由器的快速光路建立方式,该方式充分利用全光波长转换器的波长转换功能来实现虚波长双向通道,使得光通道建立的信令协议可以快速地实现网络传递,从而达到简单、方便迅捷和无阻塞的光路建立。     

自动交换光网络中恢复路径提供机制的研究

为了获得高效的网络生存性能,基于自动交换光网络(ASON)的框架,该文提出了一种新型的可恢复路径选择算法-联合可变权重可恢复路径(JVWR)选择算法,并进行了数值仿真分析,仿真结果表明,此恢复路径选择算法具有明显的业务量均衡能力,并降低了动态连接请求的阻塞概率,同时具有良好的带宽利用率和恢复资源共享效率。该文还对mesh网络业务路径和恢复路径的建立机制进行了讨论,在ASON功能框架之内,基于通用多协议标记交换提出了并行mesh共享恢复路径建立机制,从而较系统地对分布式恢复路径动态建立机制进行了研究。     

基于蚁群优化算法的网格OBS任务调度研究

本文采用蚁群优化算法（ACO）设计了基于光突发交换（OBS）网络的网格任务调度。该算法既考虑了网格用户的需求,同时结合了网格中计算资源和网络资源的分配。设计了基于ACO算法的两种通路选择规则,并在一个集中式的网格OBS仿真平台上对这两种选路规则的性能进行了对比分析。     

ASON动态光通路建立的阻塞分析

光路阻塞率是智能光网络性能重要的度量指标之一.文章对影响自动交换光网络(ASON)阻塞率的因素进行了综述,对其中的两个因素:资源保留方式和波长变换器(WC)配置方式作了重点介绍.文章还应用曾用于电路交换网的基于通路图和Lee氏的计算方法近似地对阻塞率进行了计算.     

Optical BGP Routing Convergence in Lightpath Failure of Optical Internet

Optical Border Gateway Protocol (OBGP) is an extension to BGP for Optical Cross Connects (OXCs) to automatically setup multiple direct optical lightpaths between many different autonomous domains. With OBGP, the routing component of a network may be distributed to the edge of the network while the packet classification and forwarding is done in the core. However, it is necessary to analyze the stable convergence functions of OBGP in case of lightpath failures. In this paper, we first describe the architecture of the OBGP model and analyze the potential problems of OBGP, e.g., virtual BGP router convergence behavior in the presence of lightpath failure. We then propose an OBGP convergence model derived from an inter‐AS (Autonomous System) relationship. The evaluation results show that the proposed model can be used for a stable OBGP routing policy and OBGP routing convergence under lightpath failures of the optical Internet.     

下一代光网络中光路的快速建立

文章针对下一代光网络中光路快速建立的问题,介绍了IP-over-optical的网络体系结构,并在此基础上介绍了动态RWA问题和综合解决RWA问题的动态和半动态寻路算法,讨论了这两种算法的容错性能。     

An Optical Layer Lightpath Management Protocol for WDM AONs

In this work, we propose a control protocol for lightpath management in the optical layer of all-optical networks (AONs). AONs follow a layered structure, as used by various network standards, where each layer communicates with its peer through Protocol Data Units (PDUs). In the context of Open System Interconnection (OSI) Reference Model (RM), a new layer, called optical layer, has been introduced in the AON architecture to manage the lightpath related functions. The optical layer lies in between the physical layer and the data link layer. The objective of this paper is to define a specification for an optical layer protocol for managing lightpaths in AONs. In order to study the dynamics of the optical layer protocol system, we first employ the communicating finite state machine model to represent the protocol. Then a reachability analysis of the model is performed to verify the protocol. This paper shows how the optical layer protocol can be specified formally and made error-free by the step-wise refinement of an initial specification, where validation is done after each refinement.     

基于QoS的网格工作流的选择调度算法

开放网格服务架构（OGSA）和计算经济模型的提出,使得动态的、不同QoS的服务支持下的资源调度成为一个复杂且具有挑战性的问题。文中提出了基于QoS的网格工作流的选择调度算法,该算法采用动态资源选择策略适应网格计算环境下的动态性和自治性。在满足用户对服务质量要求的基础上,平衡了工作流中服务实例的花费、处理时间和可用性之间的关系。试验证明该调度算法符合计算网格的复杂环境,能够更好地满足不同用户的实际需要。     

移动网格中依赖性任务调度

本文研究移动网格环境下的依赖性任务调度问题,提出预调度加重调度的动态调度策略.预调度利用改进的HEFT算法将所有任务映射到当前可用资源,目标是总完成时间最短,同时考虑了资源的能量约束条件.任务运行过程中监测资源的各种异常行为,当满足重调度触发条件时进行重调度,从而实现对动态环境的自适应性.通过实例分析表明了调度策略的有效性.     

移动网格可分任务调度研究

解决独立、等长度任务在性能异构且随时间动态变化的一组固定和移动资源上的调度问题,其中移动设备具有间歇连接性.提出动态循环调度DCS算法,用环结构组织任务,调度过程体现了重调度策略和复制策略.对算法进行性能分析和实验验证,表明算法性能和移动资源的间歇连接特征、任务运行跨度、移动资源数量、任务长度等因素相关.     

面向业务的弹性光网络光路损伤感知能效路由策略

针对弹性光网络中物理损伤导致业务频谱利用率低和传输能耗高问题,该文提出一种面向业务的链路损伤感知频谱分区(LI-ASP)能效路由策略。在LI-ASP策略中,为降低不同信道间非线性损伤,基于负载均衡设计一个综合考虑链路频谱状态和传输损伤的路径权重公式,根据调制方式的频谱效率和最大传输距离构造分层辅助图,从最高调制等级开始,为高质量业务选择K条边分离的最大权重传输路径;为低质量业务选择K条边分离的最短能效路径。然后,LI-ASP策略根据业务速率比值对频谱分区,采用首次命中(FF)和尾端命中(LF)联合频谱分配方式,减少不同传输速率业务间的交叉相位调制。仿真结果表明,该文所提LI-ASP策略在有效降低带宽阻塞率的同时,减少了业务传输能耗。

     

一种启发式任务调度策略

本文对分布式计算机系统中的一些任务调度策略进行了介绍和分析,进而提出和介绍了在我们的立方体分布式计算机操作系统(CDCOS)中所采用的启发式动态任务调度算法。该调度算法和CDCOS的运行都是成功的。     

A Hybrid Task Scheduling Algorithm Based on Task Clustering

Mobile Networks and Applications - Based on the problem of task communication overhead being higher than the task execution time has a direct negative impact on the makespan of task scheduling in...     

Lightpath Versus Semi-Lightpath: Some Studies on Optimal Routing in WDM Optical Networks

This paper deals with the problem of optimal dynamic routing in WDM optical networks with wavelength-changing facilities available at some of the nodes. The route may be either a lightpath (i.e., wavelength continuous channel) or a semi-lightpath (i.e., wavelength-converted channel). We attempt to estimate in this work the gain in blocking probability, when we move from lightpath routing to semi-lightpath routing, keeping the number of wavelengths fixed, in a given circuit switched network. We ensure optimal (minimum cost) routing in both the cases by using the algorithm of Banerjee et al. [7,8] (called Algorithm-I in this paper) for lightpaths and that of Chlamtac et al. [6] (called Algorithm-II) for semi-lightpaths. Our results indicate that, for both the algorithms, the blocking probability (P _B), as expected, increases with network load. At light load, P _B for Algorithm-I is always larger than that for Algorithm-II. But the rate of increase in P _B is slightly higher in case of Algorithm-II, so that there is a crossover point where P _B for Algorithm-II exceeds P _B for Algorithm-I. This probably happens due to the irregularities in the semi-lightpaths at heavy loads when almost all routes are exhausted in the network. However, since this crossover phenomenon occurs at a very congested status of the network, it has little significance over the real life operation of a network. It only suffices to indicate that, under heavy load, both the algorithms are equally insufficient, and conversion does not improve the situation as might have been expected intuitively.     

An Efficient Distributed Control Scheme for Lightpath Establishment in Dynamic WDM Networks

An adaptive hybrid reservation protocol (AHRP) is proposed for the purpose of quickly and efficiently establishing a lightpath in dynamic wavelength routed networks. This protocol uses a special reservation-and-probe (RESV_PROB) packet and extends the signaling to integrate forward reservation and backward reservation into one monolithic process. To decrease the blocking probability that happens in cases where two end nodes associated with a specific link simultaneously reserve the same wavelength, an adaptive wavelength selection policy is specially employed in AHRP. A discrete-event simulation tool based on ns-2 is developed to investigate AHRP's performance, including its blocking probability, average lightpath setup delay, and signaling overhead. AHRP is also compared with existing protocols. Results show that during highly dynamic traffic conditions, AHRP possesses the lowest blocking probability, shorter setup delay, and less signaling overhead.     

云计算环境下能量感知的任务调度算法

云计算环境下传统独立任务调度算法容易导致较高资源能耗或较大任务时间跨度.针对该问题,文中提出了两种能量感知的任务调度算法,并利用遗传算法并行化搜索合理调度方案.两种算法在搜索过程中,分别通过能耗时间归一和能耗时间双适应度方法定义适应度函数并进行个体选择.仿真结果表明,与单独考虑时间或能耗相比,这两种算法能够更有效地缩短任务执行时间跨度,降低资源能耗.