共查询到17条相似文献,搜索用时 195 毫秒
1.
移动云计算、人工智能(AI)、5G等新兴技术应用促使弹性光网络(EON)在骨干传输网中发挥更重要的角色,降级服务(DS)技术为降低EON的业务阻塞率、提高频谱利用率提供了新途径。该文首先对现有DS算法的资源分配不公、忽略低等级业务的体验质量(QoE)等问题,建立了以最小化降级频次、降级等级与传输时延损失(TDL)为联合优化目标的混合整数线性规划(MILP)模型,并提出一种时延感知的降级恢复路由与频谱分配(DDR-RSA)算法。为提高降级业务的QoE和运营商收益,在算法的最优DS窗口选择阶段中融入降级恢复策略,在保障传输数据量不变的前提下,将降级业务向空闲频域复原,从而提高频谱效率、减小降级业务TDL和最大化网络收益。最后,通过仿真证明了所提算法在业务阻塞率、网络收益和降级业务成功率等方面的优势。 相似文献
2.
弹性光网络具有高频谱效率和灵活性等优点,是光网络发展的主要方向。路由和频谱分配是弹性光网络的核心技术之一。文章提出了一种改进的基于能量感知的弹性光网络RSA(路由和频谱分配)算法,在进行选路和频谱分配时,同时考虑了网络中的能耗和可用频谱资源情况。理论分析和仿真结果表明,与传统RSA算法相比,改进算法可以提高能量效率,降低业务阻塞率。 相似文献
3.
4.
针对弹性光网络(EONs)中时变业务生存性传输时消耗频谱资源多和业务阻塞率高的问题,该文提出一种基于频谱窗滑动的时变业务共享保护(TTSP-SWS)算法。TTSP-SWS算法选择可用频谱块承载权重和保护频谱块共享度的共享保护路径代价函数值最小的保护路径;通过在保护路径上滑动频谱窗,将时变业务分配至频谱窗共享度最高的频隙位置;当时变业务带宽变化时,采用基于频谱窗滑动的生存性扩展或压缩频谱分配策略调整时变业务的带宽分配。仿真结果表明,该文所提TTSP-SWS算法能降低网络的业务阻塞率和保护资源冗余度。 相似文献
5.
为了解决虚拟光网络映射中带宽阻塞率较高以及底层资源消耗不均匀问题,论文提出一种基于时间域-频谱域碎片感知的虚拟网络映射(FA-VNM)算法。该文综合考虑频隙在时间域和频谱域上的碎片问题,设计时频联合碎片公式最小化分配过程中的频谱碎片。进一步,为了均衡网络中的资源消耗,在FA-VNM算法基础上提出基于节点度数的负载均衡感知虚拟网络映射(LB-VNM)算法,设计物理节点平均资源承载能力的公式,优先映射物理节点平均资源承载能力大的节点;为了均衡路径上资源使用,考虑路径权重值,并根据每条路径的权重值对虚拟链路进行映射,从而降低阻塞率。仿真结果表明,所提算法能有效降低阻塞率,提高资源利用率。 相似文献
6.
7.
8.
针对弹性光互联数据中心网络中任多播业务的频谱利用率低和带宽阻塞率高等问题,论文提出了一种基于光树分割和共享光路合并算法.该算法选择对光树最大长度和调制等级影响较小的树枝加入光树,使生成光树频谱代价最小.当光树传输所需频谱资源不足时,该算法将光树分割为可选较高调制等级的多子光树传输,提高频谱利用率.当网络中有足够资源时,设计一种基于链路共享度的将满足条件的子光树合并到光树上传输的机制以降低频谱消耗.仿真结果表明:论文所提算法具有较低的带宽阻塞率和较高的频谱利用率. 相似文献
9.
针对弹性光网络频谱利用率降低和带宽阻塞率高等问题,该文提出一种频谱可用性和保护带宽共享度感知的生存性多路径策略(M-SDSAA)。首先,该策略对业务采用单路径路由,通过区分业务的持续时间,分别采用首次命中和末端命中的工作路径频谱分配方式,并且为不同大小共享频谱块设定不同链路权重值。若单路径路由或频谱分配失败,引入多路径传输策略,自适应选择消耗频隙数少的策略,并且引入共享保护机制,提高保护带宽共享度。最后,当业务阻塞时,且存在单路径业务的保护路径预置在频隙消耗次优的路径,则启动重配置机制进一步降低阻塞率,将该业务重配置到保护频隙消耗最少的路径。仿真结果表明所提算法能有效地提高频谱利用率,降低网络的带宽阻塞率。 相似文献
10.
弹性光网络中多路径的保护方案相比单路径有效地降低网络带宽阻塞率,但会导致接收端多径时延差的问题,且业务的多路径分割传输策略使用了光网络较多的频谱资源。该文基于多目标遗传算法提出了遗传多路径保护算法(Genetic Multipath Protection Algorithm, GMPA),解决多路径时延差和节约频谱资源问题。在GMPA算法中,根据业务请求在光网络中建立K条边分离最短路径和带宽分配方案作为GMPA算法的初始种群,设计了一种联合考虑传输时延差和带宽资源分配的向量函数优化种群分类和拥挤距离排序。为提高算法的搜索能力和收敛速度,算法在交叉操作中设计个体自交叉方式,在变异过程中设置了带宽基因位变异范围及约束条件。仿真结果表明,相比多路径保护(Multiple Path Protection, MPP)算法和工作路径首次分配保护路径最后分配(Primary First-fit Modified Backup Last-fit, PF-MBL) 算法,GMPA算法获得最低的带宽阻塞率,其频谱资源利用率接近最优的MPP算法,路径间距离差异性能优于MPP算法。 相似文献
11.
12.
Hui Ding Min Zhang Byrav Ramamurthy Zhuo Liu Shanguo Huang Xue Chen 《Photonic Network Communications》2014,28(3):295-305
Elastic optical network is introduced as a promising technology to provide multi-bitrate-friendly data transmission in the optical layer. Elastic optical networks are based on flexible modulation format conversion, which can make more efficient use of spectrum resources than current fix-grid WDM networks. In this paper, we define the routing, modulation level and spectrum resource allocation (RMLSA) problem and then propose two novel dynamic modulation level conversion (MLC) enabled RMLSA algorithms. Numerical simulations are conducted to compare the performance of the path modulation level conversion (Path-MLC) and link modulation level conversion (Link-MLC) with different MLC thresholds with K-shortest paths RSA in terms of blocking rate, occupied spectrum and the number of MLCs across two topologies. The results verify that the Link-MLC with unlimited MLC achieves the lowest blocking rate and moderate spectrum utilization at the expense of intermediate node modulation conversions in two topologies. Smaller MLC threshold has higher blocking rate and occupied less spectrum resource. The results also suggest that the Path-MLC approach is more resource efficient than the Link-MLC with relatively higher blocking rate, and this method maybe more preferable in the networks with tight budget and/or energy constraints. 相似文献
13.
生存性虚拟光网络映射是提高光网络应对灾难故障的重要技术保障措施。为解决灾难性多区域故障导致弹性光网络的带宽容量损失问题,该文提出基于灾难预测故障模型的蚁群优化虚拟光网络映射 (DFM-ACO-VNM)算法。在该算法中,设计基于光节点资源和相邻链路的全局潜在故障概率的光节点排序映射准则,并设计启发式信息公式实现多区域故障下最小带宽容量损失的虚拟节点和虚拟链路协同映射。仿真结果表明,该文所提算法在多区域故障时能降低带宽容量损失,减少带宽阻塞率和提高频谱利用率。
相似文献14.
为了提高无色无向无冲突灵活的可重构光分插复用器(CDC-F ROADM)节点的弹性光网络IP组播频谱-能耗效率,该文提出一种全光组播能效调度算法(AMEESA)。在算法路由阶段,考虑能耗和链路频谱资源使用情况设计链路代价函数,构建最小代价光树算法组播光树。在频谱分配阶段,设计基于高效光谱分辨率(HSR)光树中间节点频谱转换方法,选择节能频谱转换方案为组播光树分配频谱块资源。仿真分析表明,所提算法能有效提升网络能效,降低IP组播带宽阻塞率。 相似文献
15.
For improving the resource efficiency of dynamic shared path protection in elastic optical networks, a survivable RSA (SRSA)-based heuristic algorithm is proposed in the paper. In SRSA, an adaptive adjustment link cost function is devised to effectively select working and protection paths. The cost function sufficiently considers available spectrum resources and the length of light paths for both working and protection paths. In order to achieve high resource efficiency, a spectrum allocation strategy named minimal cost stable set is proposed to allocate spectrum for protection paths with respect to the resource efficiency in the link cost function. And the graph coloring algorithm is introduced to select the shared protection path with the highest resource efficiency for the request. Compared with the shared path protection and dynamic load balancing shared path protection, simulation results show that the proposed SRSA decreases bandwidth blocking probability and achieves high resource efficiency. 相似文献
16.
Elastic optical network (EON) technology is considered as a very promising candidate for future high-speed networks due to its intrinsic flexibility and high efficiency in allocating the optical spectrum resources. The key issue that has to be addressed in EON is the routing and spectrum allocation (RSA) problem. RSA is NP-hard problem that has to be solved in an efficient manner. It is a highly challenging task particularly in the case of large problem instances. In this paper, we applied the bee colony optimization (BCO) metaheuristic approach to solve the RSA problem in EON with static traffic demands. The objective of the proposed BCO–RSA algorithm is to minimize both the network spectrum utilization and the average path length criterions. The results of numerous experimental studies show that our BCO–RSA algorithm performs superior compared to some benchmark greedy heuristics as well as to differential evolution (DE) metaheuristic algorithm recently proposed in the literature. The algorithm is evaluated in different realistic size optical networks, such as the NSFnet, two European optical networks (EON-19 and EON-28) and the USA network topology. Simulation results demonstrate that considerable spectrum savings could be achieved with our BCO–RSA algorithm compared to other considered approaches. In addition, we analyzed the efficiency of the BCO–RSA algorithm and compare it with the competitive DE approach according to the required CPU time and the convergence speed. 相似文献
17.