WDM网络中疏导端口的使用与分配算法研究

发布时间：2018-05-04 07:36

  本文选题：光/电混合交换 + 节能　； 参考：《西安电子科技大学》2014年硕士论文

【摘要】：随着互联网技术的迅猛发展,网络技术延伸到各个领域,通信业务量呈爆炸式增长。然而,随着人们需求的不断增加,网络带宽的不断扩大,通信设备的不断增多,能耗问题随之而来。温室效应日益恶化并且全球出现能源危机。因此降低通信设备的能耗刻不容缓。光网络的节能规划设计成为目前核心网络的研究热点之一。在光/电混合交换设备中,电交换单元的数目远远小于光交换单元的数目,一方面是由于电交换单元的功耗比较大,另一方面是由于电交换单元组成的大规模多级交换网络的成本和复杂度要高得多。但是由于在光交换网络中大量的业务传输需要业务疏导技术,所以不得不使用电交换端口。本论文的主要工作是尽量少使用端口,合理搭配使用端口和波长,在保证网络低阻塞的前提下降低网络功耗。本文中的主要工作如下:(1)提出了主节点法。根据节点的重要度大小在网络中设定主次节点,其中主节点分配电端口,具有业务疏导能力,次节点不分配电端口,没有业务疏导能力。这样大大减少了网络中使用的疏导端口数目,降低了功耗。它包括静态主节点法和动态主节点法。静态主节点法是在网络运行前已知网络负载的前提下,设定主次节点并分配疏导端口。通过仿真分析得到这种方法在低负载时节能效果特别明显,但是在负载90Erlang以后对节能没有效果。它只适合于低负载且网络负载已知的情况下。为了克服静态主节点法的缺点,又提出了动态主节点法。它能够感知网络负载的大小变化,且能根据网络负荷的变化动态地开关电端口。在负载大时及时开启一些节点的电端口以保证网络业务的正常传输,在负载小时关闭一些多余的端口从而降低功耗。仿真结果发现动态主节点法不仅在低负载时节能,在高负载时也能降低功耗还能在一定范围内降低阻塞率。(2)提出了端口分配策略。由于各个节点在网络中的疏导能力不同,同一负载下各个节点使用的疏导端口是不一样的。传统的方法是均匀分配端口,这样会造成有些节点端口不够用而有些节点出现剩余空闲端口的情况,端口得不到充分利用。端口分配策略根据节点的重要度把网络中的节点分为三类,大节点小节点和中间节点。大节点多分配端口,小节点少分配端口,中间节点分配适量的端口。本论文分别在一般环境下和动态主节点法下实施端口分配策略,仿真结果说明端口分配策略具有很好的节能效果,还能降低阻塞率,尤其是在动态主节点法下效果最好。
[Abstract]:With the rapid development of Internet technology, network technology extends to various fields, and communication traffic increases explosively. However, with the increasing of people's demand, the increasing of network bandwidth and the increasing of communication equipment, the problem of energy consumption follows. Greenhouse Effect is deteriorating and there is a global energy crisis. Therefore, it is urgent to reduce the energy consumption of communication equipment. Energy saving planning and design of optical network has become one of the research hotspots in the core network. In optical / electric hybrid switching devices, the number of electrical switching units is much smaller than the number of optical switching units, in part because of the high power consumption of the electrical switching units. On the other hand, the cost and complexity of large scale multilevel switching networks are much higher. However, a large number of traffic transmissions in optical switching networks require traffic grooming technology, so power switching ports have to be used. The main work of this paper is to reduce the power consumption of the network by using as few ports as possible, and using the port and wavelength in a reasonable way. The main work of this paper is as follows: 1) A master node method is proposed. According to the importance of nodes, the primary and secondary nodes are set in the network, in which the primary node distributes the power port, which has the ability of traffic grooming, the secondary node does not allocate the power port, and the secondary node has no traffic grooming ability. This greatly reduces the number of grooming ports used in the network and reduces power consumption. It includes static master node method and dynamic master node method. Static master node method is to set primary and secondary nodes and assign grooming ports on the premise that network load is known before network operation. The simulation results show that the energy saving effect of this method is especially obvious when the load is low, but it is not effective after the load 90Erlang. It is only suitable for low load and network load known. In order to overcome the shortcomings of static master node method, a dynamic master node method is proposed. It can perceive the change of network load and switch the electric port dynamically according to the change of network load. When the load is heavy, the power ports of some nodes are opened in time to ensure the normal transmission of network services, and some redundant ports are closed during the load hours to reduce the power consumption. The simulation results show that the dynamic master node method can not only save energy in low load, but also reduce power consumption in high load and reduce blocking rate within a certain range. Due to the different grooming ability of each node in the network, the grooming ports used by each node under the same load are different. The traditional method is to distribute the ports evenly, which will lead to some node ports are not enough and some nodes have spare ports, so the ports are not fully utilized. According to the importance of nodes, the port allocation strategy divides the nodes into three categories: large nodes, small nodes and intermediate nodes. Large nodes allocate more ports, small nodes allocate fewer ports, and intermediate nodes allocate appropriate ports. In this paper, the port allocation strategy is implemented in general environment and dynamic master node method. The simulation results show that the port allocation strategy has a good energy-saving effect, and can also reduce the blocking rate, especially in the dynamic master node method.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN929.1

【参考文献】

中国期刊全文数据库 前5条

1 张益;;一种定量评估复杂网络节点重要度的算法[J];计算机工程;2011年20期

2 李锐;李洪祚;高晓来;;OTDM(光时分复用)技术的现状和展望[J];长春师范学院学报(自然科学版);2008年12期

3 李森;;通风节能系统在电信机房中的应用[J];电信技术;2008年08期

4 王平;;对电信运营节能降耗的思考[J];电信技术;2008年07期

5 陈勇;胡爱群;胡骏;陈立全;;通信网中最重要节点的确定方法[J];高技术通讯;2004年01期

中国博士学位论文全文数据库 前1条

1 张治中;IP over WDM网络的选路和波长分配算法研究[D];电子科技大学;2002年

中国硕士学位论文全文数据库 前5条

1 薛斌;绿色协作光网络研究[D];西安电子科技大学;2014年

2 孙江峰;WDM网络中重路由算法研究[D];西安电子科技大学;2014年

3 王正平;多粒度光交换中业务疏导算法研究[D];西安电子科技大学;2012年

4 许齐钦;光网络业务疏导算法的研究[D];西安电子科技大学;2011年

5 刘晓斐;高速DWDM-PON系统性能分析[D];山东大学;2009年

，

本文编号：1842157