多种自愈环容量的比较和配置算法研究

就不同的业务需求下如何经济地设计自愈环容量进行详细探讨，其中路由配置算法是自愈环的容量设计的关键。由此，先分析了特殊业务模型下的各种自愈环容量，然后提出两种实际应用中的双向环路由配置算法－－分割法和不分割法，将它们与过去文献中所提出的平衡法（Balance Algorithm)和指派法（Demand Routing ALgorithm)进行比较，并给出模拟试验的结果，显示了这两种配置算法在充分利用SHR带宽上的优越性。     

SDH自愈环的路由算法研究

本文分析了四纤双向共享保护环的几种业务路由算法，对各种算法所需的环路容量配置进行了计算机的随机统计模拟，基于上述模拟结果，讨论了用户接入网、局间中继网等网络所具有的不同业务分布模型与路由算法的关系，探讨了根据网络自身的业务特点选择业务路由算法的方法     

ASON路由算法在电力系统中的研究

电力通信系统中的电力特种业务,特别是继电保护业务,对于通信信道的时延有着较高的要求,传统的SDH网络已经越来越不能满足其需求.本文针对业务特点,在分析了其通道延时、双向通道延时差的相关特性后,设计了一种新的ASON路由算法.仿真实验表明,该路由算法较好地实现了双向路由的寻找,并且找到的路径的时延小,算法时间复杂度低.算...     

最短路由算法在SDH复用段环网业务配置中的应用

对SDH业务路径选择等关键技术进行深入分析,总结出一种基于最短路由的环网业务配置算法。该算法在最短路由相同的情况下考虑分节点奇偶数讨论的优化方案。再通过枚举的方式进一步验证了此按节点奇偶数讨论的最短路由算法的可靠性和稳定性。     

OBS网络中基于优先级与参数可调的偏射路由算法

为了保证OBS网络中不同优先级业务的服务质量和解决偏射算法在偏射控制上的问题,提出一种基于优先级与参数可调的偏射路由算法.该算法通过设置可调参数“偏射概率”来控制冲突突发包的偏射,并从可调丢包率和偏射路径长度意义上寻找最佳偏射路由.当冲突发生时,分割优先级低的突发包,并将分割突发包偏射到空闲的链路上;在空闲的链路中选择...     

高性能安全路由器中快速路由查找算法的研究与实现

设计快速的路由查找算法是提高路由器整体性能的关键之一.文章在一种基于RAM快速路由查找算法的基础上,根据高性能安全路由器的设计要求,进一步融入Hash链式表以及Trie树查找算法的设计思想,提出了一种可配置的路由查找算法.通过动态配置算法中的评价函数系数,该算法可以适用于多种网络应用环境.     

一种片上网络自适应路由算法仿真与验证

针对片上网络的死锁问题,提出一种片上网络自适应路由算法--虚拟网络(VN)路由算法,该算法根据报文源地址和目的地址将网络分成4个虚拟网络.一旦报文在某个给定的虚拟网络中路由,所有属于最小路径的通道都可用于路由.但是,报文不能被传到另一个虚拟网络.显然,通道间没有环相关,从而避免了死锁.通过OPNET仿真,得出此算法吞吐量大、网络延迟小.     

最小费用流算法在SDH路由时隙配置中的应用

将改进的最小费用流算法应用到SDH网络路由时隙配置中。该算法自动批量配置需要上网的各种业务,并使其所需费用最小,同时网络各链路负载均衡、每条链路剩余的空闲时隙最多。在一实际的SDH网管系统中,正确性和可行性得到了验证。     

RPR多环互连自动拓扑发现研究

首先给出了实现多环自动拓扑发现的"树状链接双向链表"数据结构设计,然后讨论了基于树状链接双向链表数据结构的弹性分组环多环网络中顺时针分层自动拓扑发现算法的实现流程,并详细分析了实现多环自动拓扑发现时的Layer 2路由算法,给出了该算法实现的一个实例,最后结合该实例给出了多环自动拓扑发现算法的仿真研究结果。     

基于MED属性的全网iBGP配置的路由预测算法

当网络经营者为了某种目的修改网络配置时,必须预测配置改变所带来的潜在危险.给出了一种算法,该算法不需要对路由协议动态进行复杂的模拟,只需要静态的路由数据就可以计算出全网iBGP配置的路由选择结果,并在一个模拟的大型的ISP上验证了预测算法的正确性.     

针对MPLS网络流量工程的链路关键性路由算法

该文针对多协议标签交换(MPLS)网络流量工程提出了一种链路关键性路由算法(LCRA),该算法通过定义链路的平均期望负载来确定链路的关键性,进而映射为链路的权值来决定路由的选择。该算法的目的是使网络快捷地建立尽可能多的有带宽保证的路由,并且使这些路由均衡通过网络,实现网络负载均衡。与其他算法相比,该算法在路由拒绝率和重路由性能方面有很好的表现,并且路由的建立时间相当快捷。     

DraLCD:一种新的数据中心流量工程方法

流量均衡是为了避免网络拥塞而作为流量工程中的路由优化目标提出来的,由于数据中心网络的流量特性,使得传统IP网络的流量工程方法不一定适合.为此,本文在SDN(Software Defined Network)的框架下,提出了一种基于链路关键度的自适应负载均衡流量工程方法:DraLCD(Dynamic Routing Algorithm based on Link Critical Degree).该方法通过对全局视图的网络管控,并充分利用了网络中存在的冗余路径,在完成细粒度流量均衡的同时,能够降低控制器的计算开销以及与交换机之间的通信开销,最终完成路由优化的目标.最后,基于DraLCD设计的原型系统,通过在Mininet仿真平台中部署并进行仿真实验,与现有的等开销多路径路由算法ECMP(Equal-Cost Multi-Path)以及GFF(Global First Fit)路由算法相比较,能够明显地提升网络性能.     

Modified Least Load Multicast Routings for Single Rate Fully-Connected Loss Networks

In this paper, we investigate the dynamic multicast routing problem and briefly discuss the common dynamic multicast routing algorithm called Least Load Multicast Routing (LLMR). We propose two new multicast routing algorithms called Least Load Multicast Routing with Maximum Occupied Circuits (LLMRMOC) and Least Load Multicast Routing with Minimum Measured Blocking Time (LLMRMMBT), both of which are based on LLMR. Simulation results show that these algorithms, compared with LLMR, not only produce a smaller network revenue loss, but also result in smaller call blocking probabilities for all classes of traffic. The moderate gain in the network performance comes only a slight additional cost. The implementation issues of the proposed algorithms are also discussed.     

Clos网络中的组播路由算法

对于三级Clos网络,扇出机制会影响Clos网络的阻塞率、算法的时间复杂度及网络成本,因此选择好的扇出方式能充分发挥网络的组播能力。根据输出级扇出、中间级扇出、输入级扇出等不同的扇出机制分类,可将组播算法分为输入级扇出算法（IFMA）、最迟扇出算法（LFMA）、切割扇出算法（SFMA）、中间级优先扇出算法（CMFF—MA）。在对4种算法仿真比较的基础上,文章提出针对不同的业务采用不同的处理方法的路由方案,对于固定扇出业务可采用CMFFMA算法进行路由,针对递增业务采用先输出级、再中间级、最后输入级扇出的策略,可有效地降低阻塞率。     

专用网路由方法的改进及抗毁策略

饱和路由方法是Ｅｕｒｏｃｏｍ推荐的一种路由方法,由于其高可靠性,在分组交换网中得到广泛应用。但是,其固有弱点在于泛洪式广泛信息往往不可避免地带来网络中过多的信息冗余,从而导致网络性能下降。本文在保证网络可靠性的前提下,提出了直达路由与抗毁相结合的策略。理论分析与计算机仿真表明,此策略可以明显地改善网络性能,同时不影响其鲁棒性。     

SDH多业务传输平台(MSTP)路由算法

多业务传输平台(MSTP)的出现使得同步数字体系(SDH)网络可以传输多种业务.路由和保护模块是网络管理软件的核心模块之一.由于网络成本的限制,在网络中往往会出现多种工作方式不完全一致的设备,增加了路由的复杂性.文章对多设备多业务的SDH网络的路由和保护进行了分析,分别给出了电路业务和数据业务的路由算法.此算法已成为华环公司网络管理系统的核心模块.     

Distributed multicast multichannel paths

Supporting multimedia applications in QoS-aware multicast deployment has become an important research dimension in recent years. Future communication networks will face an increase in traffic driven by multimedia applications with stringent requirements in the following important functions: (1) nodes and links used distributing, (2) packets duplication distributing, (3) QoS supporting, (4) multichannel routing. For improving these four functions, in this paper we propose a new polynomial time algorithm, named Nodes Links Distributed-Multicast Multichannel Routing (NLD-MMR), based on the Constraint-Based Routing (CBR) and Linear Programming (LP). The new algorithm by constructing Distributed Multicast Multichannel Paths (DMMCP) can distribute or compact both paths and traffic. Our simulation study shows that the proposed algorithm, as compared to other available algorithms, performs well and constructs a new generation of optimal paths with the best cost and efficiency.     

New Approaches to Routing in Mobile Ad hoc Networks

Worldwide Interoperability for Microwave Access (Wimax) is power station through which mobile network, commonly known as A Mobile Ad-hoc Network (MANET) is used by the people. A MANET can be described as an infrastructure-less and self-configure network with autonomous nodes. Participated nodes in MANETs move through the network constantly causing frequent topology changes. Designing suitable routing protocols to handle the dynamic topology changes in MANETs can enhance the performance of the network. In this regard, this paper proposes four algorithms for the routing problem in MANETs. First, we propose a new method called Classical Logic-based Routing Algorithm for the routing problem in MANETs. Second is a routing algorithm named Fuzzy Logic-based Routing Algorithm (FLRA). Third, a Reinforcement Learning-based Routing Algorithm is proposed to construct optimal paths in MANETs. Finally, a fuzzy logic-based method is accompanied with reinforcement learning to mitigate existing problems in FLRA. This algorithm is called Reinforcement Learning and Fuzzy Logic-based (RLFLRA) Routing Algorithm. Our proposed approaches can be deployed in dynamic environments and take four important fuzzy variables such as available bandwidth, residual energy, mobility speed, and hop-count into consideration. Simulation results depict that learning process has a great impact on network performance and RLFLRA outperforms other proposed algorithms in terms of throughput, route discovery time, packet delivery ratio, network access delay, and hop-count.

     

CPS异构分布实时系统中带负载因子的容错调度

介绍了Cyber-Physical Systems的基本内容,给出了基于Cyber-Physical Systems异构分布式中的实时任务系统模型。并在该模型下结合基/副版本备份技术提出了两种适应于Cyber-Physical Systems异构分布式实时环境的启发式容错调度算法:HDLMA算法和HDLFA算法。最后针对这两种算法,分析了算法可调度性,负载均衡性,任务粒度大小对负载均衡性的影响,以及调度阀值对算法可调度性的影响。     

Q-MOHRA: QoS Assured Multi-objective Hybrid Routing Algorithm for Heterogeneous WSN

Quality of Service (QoS) assurance in Wireless Sensor Network (WSN) is a tough task, and it is more exciting due to the scarcity of resources. The requirement of different WSN applications running over has different constraints. In QoS, routing protocol the network has to balance the traffic. This paper presents a novel heuristic routing algorithm known as QoS assured Multi-objective Hybrid Routing Algorithm (Q-MOHRA) for Heterogeneous WSN. Q-MOHRA takes into account the link (energy, hop count, link quality indicator etc.) and path (jitter) metrics for optimal path selection. The performance of Q-MOHRA is evaluated through intensive simulation and equated with Simple Hybrid Routing Protocol (SHRP) and Dynamic Multi-objective Routing Algorithm (DyMORA). The metrics such as average energy consumption, residual energy, packet delivery ratio, jitter, and normalized routing load are used for comparison. The performance of Q-MOHRA has been observed to outclass SHRP and DyMORA. It improves the packet delivery ratio by 24.31% as compared to SHRP and 11.86% as compared to DyMORA. Q-MOHRA outperforms DyMORA in terms of average energy consumption by a factor of 8.27%.