静态波带交换中的一种波带粒度取值算法

波带交换可以有效地减少波长交换的交换端口数量,本文研究节点间业务量已知时静态波带交换中的波带粒度取值算法,提出了基于k均值聚类的波带粒度取值算法。算法将业务量相近的业务分为一组,一组内的业务用相同粒度的波带装载,以提高波带的利用率。研究表明,在没有业务疏导的环境下,与其他方法相比,算法使用的波带数量和波带内的空闲波长数量都比较少。本文还研究了静态疏导环境下不同波带粒度取值算法的性能,提出了多波带粒度下的业务装载策略。对于大粒度的波带,使用向下装载,而对于小粒度的波带,使用向上装载,意在减少波带的使用数量的同时提高波带利用率。仿真结果表明,使用静态业务疏导后,本文算法与其他方法相比,依然可以有效地减少波带数量,提高波带利用率。与基于组播路由的静态波带疏导算法相结合,使波带利用率可以达到98%以上。     

一种基于Credit的变长分组并行交换网络调度算法

该文提出了一种新的并行分组交换(PPS)网络调度算法。该算法通过在解复用器处采用以变长分组为业务分配单元的方式消除了信元的乱序问题;通过采用Credit机制进行业务分配,实现了业务到各个交换平面完全公平的分配;各个并行交换单元采用组合输入输出排队,降低了对缓存和交换平面的加速要求,同时可以充分利用现有单Crossbar网络调度算法的研究成果。文中证明了该算法对业务分配的公平性,对高速缓存的需求量以及整个网络的稳定性,仿真进一步证明了该算法具有良好性能。     

一种基于虚拟突发的OBS网络业务疏导机制

提出了一种基于虚拟突发光突发交换的(OBS)网络业务疏导机制TG-VB(traffic grooming based on virtual burst)。作为一种多粒度动态业务疏导机制,TG-VB将常规突发通过"逻辑汇聚"组成虚拟突发,增大核心网中业务传输粒度;同时,TG-VB根据虚拟突发整体传输、整体调度的特点,为汇聚的常规突发动态地选择最优的共用传输路径。仿真结果表明,与采用传统疏导机制的OBS网络相比,TG-VB能够有效地减轻核心节点的通信处理开销,降低对网络节点端口数量的需求;在相同负载条件下,TG-VB可极大地提高波长信道利用率。     

基于可调光滤波器的波带交换节点性能研究

提出了一种基于可调光滤波器的波带交换节点结构,讨论了这种节点在不同波带算法下的性能.结果表明,这种波带交换节点结构能有效降低波带交换节点端口数目,提高网络吞吐量,具有很好的网络性能.     

波带交换光网络不同路由策略下端口使用效率

研究了在不同波带组合和波带交换算法下渡带交换网络的性能,结果表明,使用波带交换为业务请求建立连接可更有效地节约网络资源,降低网络管理的复杂度.     

波带交换新算法及其性能分析

由于波带交换在减少交换结构端口数、降低控制面复杂度和光交叉连接费用方面具有实际的意义,因而引起了人们广泛的关注.文章提出了两种新的波带交换算法并且做了相应的性能分析.     

自适应混合光交换

当前的混合光交换只是简单地把光电路交换和光突发交换叠加在一起,其复杂度高.并且不能有效支持突发性强的业务.针对这种缺陷,本文提出了一种自适应混合光交换方式.自适应混合光交换的入口边缘节点监测一对入口边缘节点和出口边缘节点之间的业务流量的大小,自适应地调整光交换方式.在最小突发长度最大汇聚时间算法的基础上,提出了自适应混合光交换的汇聚算法,以提高光路的利用率和有效支持突发性强的业务.同时提出了数据信道调度算法,它能统一为光路和数据突发预留资源,并降低了混合光交换的复杂度,提高了网络性能.仿真结果表明,与光突发交换相比,自适应光交换能有效降低IP包丢失率和端对端时延.     

多粒度光网络波带组合策略的研究

为了有效地传输不同波长粒度的业务,对多粒度光网络波带组合进行了研究。首先介绍了多粒度光网络的模型和光交叉节点,并在此基础上分析了多粒度交叉连接结构的两种结构,单层和多层多粒度交换结构。结合多粒度的交换结构,给出了几种波带组合策略,并对这些策略进行了分析和比较。     

业务疏导技术在WDM网络及OPS网络中的应用

在波分复用(WDM)光网络中,可使用业务疏导(Traffic Grooming)技术来提高网络性能,降低网络成本.详细阐述了WDM光网络中业务疏导的基本概念及主要目标,并对国内外研究现状进行了总结.最后介绍了OPS光交换网络中使用的业务疏导技术.     

光突发交换宽带多业务接入技术探索

文章探索了一种以低廉价格实现多业务高速接入的光突发交换(OBS)接入网方案,方案采用OBS技术实现多业务传送平台.文章介绍了该方案的网络架构、边缘节点和核心节点的结构,以及基于多协议标签交换(MPLS)的控制平面和数据平面的软件架构;重点研究了一种在边缘节点处权衡带宽利用率和实时业务延时要求的业务分类和调度机制,利用该机制可以实现满足不同业务要求的多业务高速接入,并可根据实时性业务低延时的要求来合理地分类和设置最大等待时间.     

一种新型光交换网络的流量特性分析和传输机制研究

阐述了可以弥补光电路交换和光突发交换二者缺点的混合交换技术,提出了一种采用混合交换技术的新型光交换网络,分析了该新型网络中的流量特性并提出了与该网络相应的传输机制.实验表明,采用该传输机制的新型光交换网络,在源负载低于0.75的情况下,其网络丢包率和吞吐量等性能均优于普通混合交换网络,能达到优化网络资源的目的.     

基于连通支配集的OBS网络多粒度业务疏导机制

针对目前光网络核心节点的交换速度无法匹配单波长的传输速度难以满足大量的业务请求而拥塞,为了提高核心节点交换效率使其具有大容量数据交换的能力,本文提出了一种基于连通支配集(CDS)的光突发交换(OBS)网络稀疏节点疏导机制(CDS-TG).主要思路是:首先在OBS核心网络中根据改进的连通支配集算法选取疏导节点;其次在疏导...     

Effective Traffic Grooming Algorithms in SONET/WDM Ring Networks

Much work has focused on traffic grooming in SONET/WDM ring networks. Previous work has considered many aspects of traffic grooming, including minimizing the number of ADMs, minimizing the number of wavelengths, considering different traffic models, using different network architectures, incorporating switching capability and so on. In this work, we study traffic grooming in unidirectional ring networks with no switching capability under both uniform traffic and non-uniform traffic models to reduce electronic multiplexing costs. Based on the clustering notion, we derive a general and tighter lower bound for the number of ADMs required in traffic grooming under the uniform all-to-all traffic model. This bound reduces to special cases obtained in previous work. We also derive general, tighter, and closed form lower bounds for the number of ADMs required under two non-uniform traffic models: the distance-dependent traffic model and the non-uniform symmetric traffic model. Cost-effective multi-phase algorithms that exploit traffic characteristics are then designed and studied to efficiently groom traffic streams under different traffic models. Our numerical and simulation results show that the proposed multi-phase algorithms outperform existing traffic grooming algorithms by using a fewer number of ADMs. Our algorithms in several cases also achieve the lower bounds derived.     

Green provisioning of the traffic partition grooming in robust,reconfigurable and heterogeneous optical networks

In recent years, various high-speed network architectures have been widespread deployed. Dense Wavelength Division Multiplexing (DWDM) has gained favor as a terabit solution. The optical circuit switching has also been provided for “sub-rate” aggregation. Such that, the granular types of demands tend to be diverse and must be evaluated. However, current dedicated optical networks do not offer sufficient flexibility to satisfy the requirements of demands with such wide range of granularities. The traffic grooming becomes a power-efficient one only when it does not utilize the aggregation of Coarse-Granularity (CG) demands. The waveband switching merely provides port-cost-effective connections for CG demands regardless of fine-granularity ones. Consequently, in this paper, we devise a heterogeneous grooming method called traffic partition grooming. It combines the power efficiency advantage of the traffic grooming under fine-granularity environment and the port savings advantage of the waveband switching under coarse-granularity environment to provide green provisioning. In addition, the optical virtual topology self-reconfigures along with various optimization objectives variation and has the robustness to determine the pre-unknown information. This paper is also the first work on investigating the issue of Robust, Reconfigurable and Heterogeneous Optical Networking (R²HON). The effective green provisioning and OPEX savings of our R²HON have been demonstrated by numerical simulations.     

Multicast multi-granular grooming based on integrated auxiliary grooming graph in optical networks

In order to improve the bandwidth utilization efficiency and reduce the blocking probability for multicast requests in optical networks, the multicast traffic grooming approach was proposed to groom a lot of low-speed traffic to a few of high-speed light-trees. At the same time, in order to save the ports and the cost of optical cross-connect, the multicast waveband grooming approach was proposed to groom multiple light-trees to a few of waveband tunnels. However, the existing approaches all did not consider the joint performances of improving bandwidth utilization efficiency, reducing blocking probability and saving ports for multicast requests. Therefore, in this article, we propose a new multicast multi-granular grooming approach to perform the hierarchical sequential grooming to improve the joint performances based on a newly developed integrated auxiliary grooming graph (IAGG) that includes multiple wavelength integrated grooming graphs (WIGGs) and one waveband virtual topology layered graph (BVLG) to support both the multicast traffic grooming and multicast waveband grooming. In order to achieve the map of light-tree to the virtual topology layer in WIGG or BVLG, we present a light-segment map method, where a light-tree will be divided to several light-segments each of which will be independently mapped to the virtual topology layer. Since different definitions of blocking probability may lead to different objectives, we define two kinds of blocking probability, mean blocking probability of requests (MBPR) and mean blocking probability of users (MBPU). According to the two definitions of blocking probability, we propose two multicast multi-granular grooming heuristic algorithms, Heuristic Algorithm with minimizing MBPR and Heuristic Algorithm with minimizing MBPU based on IAGG. Simulation results show that the two proposed algorithms are both efficient and have better performances than traditional multicast grooming algorithm.     

Survivable integrated grooming in multi-granularity optical networks

Survivability is an important issue to ensure the service continuity in optical network. At the same time, with the granularity of traffic demands ranging from sub-wavelength-level to wavelength-level, traffic demands need to be aggregated and carried over the network in order to utilize resources effectively. Therefore, multi-granularity grooming is proposed to save the cost and reduce the number of switching ports in Optical-Cross Connects (OXCs). However, current works mostly addressed the survivable wavelength or waveband grooming. Therefore, in this paper, we propose three heuristic algorithms called Multi-granularity Dedicated Protection Grooming (MDPG), Multi-granularity Shared Protection Grooming (MSPG) and Multi-granularity Mixed Protection Grooming (MMPG), respectively. All of them are performed based on the Survivable Multi-granularity Integrated Auxiliary Graph (SMIAG) that includes one Wavelength Integrated Auxiliary Graph (WIAG) for wavelength protection and one waveBand Integrated Auxiliary Graph (BIAG) for waveband protection. Numerical results show that MMPG has the lowest average port-cost, the best resource utilization ratio and the lowest blocking probability among these three algorithms. Compared with MDPG, MSPG has lower average port-cost, better resource utilization ratio and lower blocking probability.     

Waveband Switching for Dynamic Traffic Demands in Multigranular Optical Networks

Waveband switching (WBS) in conjunction with multigranular optical cross-connect (MG-OXC) architectures can reduce the cost and complexity of OXCs. In this paper, we study the performance of different MG-OXC architectures under dynamic traffic. In the case with online incremental traffic, we compare two MG-OXC architectures in terms of the blocking probability of new lightpath requests and study the impact of port counts and traffic loads. We develop an online integer linear programming model (On-ILP), which minimizes the number of used ports and the request blocking probability, given a fixed number of wavelengths and MG-OXC architecture. The On-ILP optimizes the routing of new lightpaths so as to maximize lightpath grouping and reduce the port count given that existing traffic cannot be rearranged. We also propose a new efficient heuristic algorithm, called maximum overlap ratio (MOR) to satisfy incremental traffic and compare it with the On-ILP, first-fit, and random-fit algorithms. Our results and analysis indicate that using WBS with MG-OXCs can reduce the size (and, hence, the cost) of switching fabrics compared to using ordinary OXCs. Based on the results and observations in the incremental traffic case, we further study the performance of a particular MG-OXC architecture under fully dynamic or fluctuating traffic. Our simulations show that the proposed heuristic algorithm waveband assignment with path graph, which groups wavelengths to bands and uses wavelength converters efficiently under fluctuating traffic, significantly outperforms other heuristic algorithms.     

Waveband Grooming based on Layered Auxiliary Graph in multi-domain optical networks

With the number of wavelengths on fibers keeps increasing, the size and the cost of Optical Cross-Connect (OXC) are greatly enhanced and then the control and management of optical switches become more and more complicated. Therefore, the technique called waveband switching is proposed to reduce the size and the cost of OXC; that is, to save the All-Optical (OOO) switching ports in OXC. However, the existing waveband switching algorithms are all limited in single-domain optical networks. Actually, with the scale of optical backbone keeps enlarging, the network is divided to multiple independent domains to perform the hierarchy routing for achieving the scalability. In order to reduce the size and the cost of OXC meanwhile to achieve the scalability in multi-domains, in this paper we propose a new heuristic algorithm called Waveband Grooming with Layered Auxiliary Graph (WGLAG) since the waveband grooming problem is the NP-hard to perform the inter-domain routing based on the virtual topology of multi-domain network and the intra-domain routing based on the physical topology of single-domain network. In intra-domain routing with waveband grooming of each single-domain, we propose the Layered Auxiliary Graph (LAG) that includes one virtual topology layer and multiple waveband-plane layers to compute a single-hop, or multi-hop or hybrid waveband route for each connection request based on the sub-path waveband grooming scheme. Simulation results show that, WGLAG not only can effectively save more switching ports to reduce the cost of OXC but also can obtain lower blocking probability than other algorithm.