基于时延估计组合模型的NCS控制方法研究

在对网络传输时延的组成和特性进行分析的基础上,针对双边网络控制系统,提出了一种时延估计组合模型,并对该组合模型的有效性进行了评估。利用该模型进行网络时延估计,估计结果用来设计网络时延补偿控制器,以实际网络时延为实验数据进行了仿真实验,实验结果表明网络控制系统的动态性能有了明显改善,同时进一步证实了时延估计组合模型的有效性。     

基于动态迁移的光传送网SDN低时延调度研究

光传送网SDN调度较传统网络具有更好的动态性,为了使SDN网络性能与大数据和云计算需求匹配,提出了一种低时延动态迁移策略.先以SDN网络结构、流表耗时、以及时延调整系数等变量构建通信代价模型,得到通信代价的目标与约束.基于通信代价模型对SDN网络架构采取动态迁移,根据路径请求量计算控制器的平均时延与累计效用,再利用控制与数据信道负载对流表采取动态调整.通过仿真,确定了光传送网SDN调度过程中,当控制器个数改变时,时延几乎不受影响;当数据量增加时,时延及其增速也能得到有效控制;在动态迁移后,负载差异度降低至迁移前的71.92％.实验结果表明所提方法能够有效降低光传送网SDN调度时延,并保证良好的网络负载均衡.     

基于软件定义网络的边缘控制部署机制

针对软件定义网络（software-defined Networking,SDN）中单一控制器容易发生过载导致较长时延的问题,提出一种基于SDN的边缘控制模型,该模型采取分层部署方式将边缘计算集成到SDN中,每个边缘控制器控制其覆盖范围内部署的所有子边缘控制器和交换机,负责区域内网络设备的通信量。为了方便管理边缘控制器之间的交互,该模型引入一个控制器代理模块,将设备请求转发给父控制器或将路由信息发送给子控制器来协调控制器之间的工作。实验结果表明,与基于SDN的传统网络相比,该方法依托部署在网络设备边缘的计算和存储服务,减轻了SDN主控制器上的负载,降低了转发平面和控制平面之间的延时,显著地改善了总处理延时和带宽使用情况。     

针对大规模软件定义网络的子域划分及控制器部署方法

针对大规模软件定义网络（SDN）的多控制器部署模型计算复杂度高的问题,定义了控制链路可靠性等多个衡量网络服务质量的指标,并提出一种针对大规模SDN的子域划分及控制器部署方法。首先,该方法利用改进的标签传播算法（LPA）将网络划分成多个子域,然后在子域中分别部署控制器。在考虑控制链路平均时延、可靠性以及控制器负载均衡等多个性能指标的基础上,将问题模型的计算复杂度降低至仅与网络规模呈线性关系。实验结果表明,所提算法与原始的LPA相比,控制器负载均衡性得到明显优化;与容量受限的控制器部署（CCP）算法相比,模型的计算复杂度和网络服务质量得到明显改善：在Internet2拓扑中,控制链路平均时延最多减小9%,控制链路可靠性最多增强10%。     

基于改进樽海鞘群算法的SDN控制器部署算法

针对SDN(Software-Defined Networking)多控制器部署优化问题,提出一种基于改进樽海鞘群算法的SDN控制器部署算法.在元启发式算法樽海鞘群优化的基础上,通过引入混沌映射因子增加收敛性,避免优化器陷入局部最优,提高算法性能,从而可以更有效地动态评估大型SDN网络中控制器的最佳数量以及交换机和控制器之间的最佳连接.实验结果表明,该算法在执行时间和可靠性上均优于其他对比算法,实现了控制器的最佳数量以及控制器与交换机的最佳分配问题.     

SDN中基于遗传机制的自适应路由算法研究

SDN以集中式的控制、可编程的接口等优点,极大地提高了网络的管控效率及操作的灵活性。但在SDN部署运行中,也暴露出传输时延大、丢包率高等缺点。针对这些问题,提出了一种基于遗传算法的自适应SDN路由算法,该算法利用遗传算法在SDN的全局网络视图中搜索优化路径。算法设计时,对交叉、变异操作进行条件约束,避免产生无效的路径,减小求解空间,降低控制器计算开销。同时能根据网络的动态变化,自适应地选择转发路径。通过Mininet仿真平台进行实验,与其他算法相比,该算法降低了网络的时延以及丢包率,提高了网络性能。     

基于极点配置的随机时延网络控制器设计

针对网络控制系统(NCS)的随机短时延问题,在传感器和执行器由时间驱动、离散控制器由事件驱动方式下,将NCS描述为切换系统模型。结合极点配置和特征根连续变化原理,根据时延与增广矩阵特征根最大模长之间的关系,将切换系统进行简化并设计相应的状态反馈控制器。实验结果表明,该控制器求解简单、易于实现,且可根据实际系统性能要求选择控制参数。     

基于SDN的数据中心动态优先级多路径调度算法

随着云计算技术和分布式业务的发展,数据中心内部“东西向”大象流量激增,这部分大象流在调度不当的情况下容易发生碰撞,造成链路拥塞。本文提出一种基于软件定义网络（SDN）的动态优先级多路径调度算法（DPMS）。该算法根据数据中心流量的特点制定大象流和老鼠流调度模型,充分利用各网络节点间的冗余链路,提高资源利用率;并结合组表优化SDN架构中控制器和交换机的通信模式,降低了数据包处理时延。实验结果表明,相比ECMP和Hedera这2种调度策略,DPMS提高了网络吞吐量和链路利用率,减少了平均流完成时间,网络的整体性能有所提高。     

SDN控制器部署中的可靠性优化研究

软件定义网络(SDN)将传统网络结构中的控制层和转发层解耦,其将所有转发设备与一个逻辑集中的控制器相连接。为避免网络规模不断扩大引起的单点失效,向分布式控制结构发展成为Open Flow广域网部署的趋势,其中控制层多控制器的部署问题是SDN设计中的一个关键环节。提出基于控制路径连通度的控制器部署方案来最大化SDN控制器部署的可靠性,并使用3种不同的算法对比控制器部署性能。仿真结果表明,该方案可以在可接受时延范围内提升部署SDN控制器的可靠性。     

基于多模电台的SDMANET网络组网方法

软件定义网络（SDN）在有线网络和数据中心网络等多种网络场景中正在快速发展,然而在移动自组织网络（MANET）中SDN的使用仍然处于起步阶段。因MANET网络拓扑变化频繁、资源受限以及采用分布式组网方式等特点使得在其中应用SDN变得具有挑战性。为此,本文提出一种基于多模电台的软件定义移动自组织网络（SDMANET）组网方法。该方法使用支配集算法计算骨干节点,仅由骨干节点使用带外信道和SDN控制器通信,并在MAC层基于骨干节点进行TDMA时隙动态分配。实验结果表明,与OLSR协议和直接带外控制的SDN方法相比,本方法具有更低的网络控制开销和信道访问时延,在大规模MANET网络中性能较好。     

Comparison of PI controllers designed for the delay model of TCP/AQM networks

One of the major problems of communication networks is congestion. In order to address this problem in TCP/IP networks, Active Queue Management (AQM) scheme is recommended. AQM aims to minimize the congestion by regulating the average queue size at the routers. To improve upon AQM, recently, several feedback control approaches were proposed. Among these approaches, PI controllers are gaining attention because of their simplicity and ease of implementation. In this paper, by utilizing the fluid-flow model of TCP networks, we study the PI controllers designed for TCP/AQM. We compare these controllers by first analyzing their robustness and fragility. Then, we implement these controllers in ns-2 platform and conduct simulation experiments to compare their performances in terms of queue length. Taken together, our results provide a guideline for choosing a PI controller for AQM given specific performance requirements.     

A robust fractional-order PID controller design based on active queue management for TCP network

In this paper, a robust fractional-order controller is designed to control the congestion in transmission control protocol (TCP) networks with time-varying parameters. Fractional controllers can increase the stability and robustness. Regardless of advantages of fractional controllers, they are still not common in congestion control in TCP networks. The network parameters are time-varying, so the robust stability is important in congestion controller design. Therefore, we focused on the robust controller design. The fractional PID controller is developed based on active queue management (AQM). D-partition technique is used. The most important property of designed controller is the robustness to the time-varying parameters of the TCP network. The vertex quasi-polynomials of the closed-loop characteristic equation are obtained, and the stability boundaries are calculated for each vertex quasi-polynomial. The intersection of all stability regions is insensitive to network parameter variations, and results in robust stability of TCP/AQM system. NS-2 simulations show that the proposed algorithm provides a stable queue length. Moreover, simulations show smaller oscillations of the queue length and less packet drop probability for FPID compared to PI and PID controllers. We can conclude from NS-2 simulations that the average packet loss probability variations are negligible when the network parameters change.     

Adaptive generalized minimum variance congestion controller for dynamic TCP/AQM networks

Generic generalized minimum variance-based (GMV) controllers have been adopted as efficient control mechanisms especially in presence of measurement noise. However, such controllers exhibit degraded performance with change in process dynamics. To overcome this problem, a novel congestion controller based on active queue management (AQM) strategy for dynamically varying TCP/AQM networks known as adaptive generalized minimum variance (AGMV) is proposed. AGMV is the combination of the real-time parameter estimation and GMV. The performance of the proposed scheme is evaluated and compared with its adaptive minimum variance (AMV) counterpart under two distinct scenarios: TCP network with unknown parameters and TCP network with time varying parameters. Simulation results indicate that, in either case, AGMV is able to keep the queue length around the desired point. In addition, the superior performance of the proposed controller has been shown with regard to the PI controller, which is well-known in the AQM domain.     

分布式SDN控制器放置问题研究

软件定义网络(Software-Defined Networking,SDN)通过控制平面与数据平面的分离和逻辑集中的控制构建了新的网络范式.考虑性能、可扩展性和可靠性等方面的需求,大规模网络通常采用分布式SDN控制平面,即通过放置多个控制器共同管理整个网络.这需要确定控制器的放置数量、放置位置以及交换机到控制器的分配...     

Resource Allocation for Reliable Communication Between Controllers and Switches in SDN

In software-defined networking (SDN), the communication between controllers and switches is very important, for switch can only work by relying on flow tables received from its controller. Therefore, how to ensure the reliability of the communication between controllers and switches is a key problem in SDN. In this paper, we study this problem from two aspects: the controller placement and the resource backup aspect. Firstly, in order to implement the reliable communication and meet the required propagation delay between controllers and switches, a min-cover based controller placement approach is proposed. Then, in order to protect both controllers and control paths from regional failure, a backup method based on an exponential decay failure model is proposed, which considers the regional influence and the survivability of backup controllers and control paths. Simulations show that our controller placement approach can meet the reliability and delay requirement with appropriate controller allocation scheme, and our backup method can improve the survivability of backup controllers and control paths while ensuring the performance of control network.     

MAP-SDN: a metaheuristic assignment and provisioning SDN framework for cloud datacenters

Software-defined networking (SDN) introduces a new method in networking that by offering programmability and centralization, it can dynamically control and configure networks. In traditional networks, data plane did the whole forwarding process, but SDN decouples data plane and control plane by using programmable software controllers for deciding how to forward different flows. By implementing control plane in a software-based independent layer, the network management will become much easier and new policies can be applied to the network by changing a few lines of code. Since the resource allocation and meeting the required service-level agreement are really important in large-scale networks such as cloud datacenters, using SDN can be very useful. In these networks, one logically centralized controller cannot handle the whole network traffic and it will become network bottleneck. Therefore, multiple distributed controllers should be allocated in different regions of the network. Since the request rate of switches varies in time, by dynamic allocation of controllers, network resources will be allocated efficiently and this approach can also reduce power consumption. In this paper, we are going to propose a framework for provisioning software controllers in cloud datacenters by using metaheuristic algorithms. These algorithms can be less accurate compared to other kinds, but their main characteristics like simplicity, flexibility, derivation free, and local optimum avoidance make them a good nominee for solving controller provisioning problem and controller placement problem. Our framework improves computation time and reaches better results compared to other allocation techniques, but it is less accurate in some scenarios. Therefore, we believe metaheuristic approach can be very useful in developing new technologies for SDN in the future.     

Design of robust active queue management controllers for a class of TCP communication networks

This paper describes the design of active queue management (AQM) controllers for a class of TCP communication networks. In TCP/IP networks, the packet-dropping probability function is considered as a control input. Therefore, a TCP AQM controller was modeled as a time-delayed system with a saturated input. The objective of the work described here was to design robust controllers capable of achieving the desired queue size and guaranteeing asymptotic stability of the operating point. To achieve this aim, we have proposed two control strategies, namely a static state feedback controller and an observer-based controller. By applying the Lyapunov-Krasovskii functional approach and the linear matrix inequality technique, control laws and delay-independent stability criteria for the AQM controllers were derived. The performance of the two control schemes was evaluated in various network scenarios via a series of numerical simulations. The simulation results confirm that the proposed schemes outperform other AQM schemes.     

面向智慧医疗云的SDN动态负载均衡方法

文中 引入软件定义网络(Software Defined Network,SDN)对智慧医疗云进行网络管理,并且针对传统SDN控制器存在单点失效和负载均衡的问题,设计了智慧医疗分布式SDN控制器系统。SDN控制系统分为SDN控制器集群、数据转发平面和智慧医疗云服务系统3层。在此基础上,提出一种实时负载动态自调节的快速负载均衡算法DAF(Dynamic Adaptive and Fast Load Balancing)。在该算法中,负载信息感知组件周期性地采集自己的负载信息,自动地进行控制器间的负载信息交互;控制器的负载值超过阈值时,会触发交换机迁移动作,以动态配置交换机与控制器之间的映射关系。实验结果表明,面向智慧医疗云的分布式SDN控制系统的性能良好,且DAF算法能够快速地实现SDN控制器间的负载均衡,提升了智慧医疗云的网络吞吐量。     

A <Emphasis Type="Italic">K</Emphasis> self-adaptive SDN controller placement for wide area networks

As a novel architecture, software-defined networking (SDN) is viewed as the key technology of future networking. The core idea of SDN is to decouple the control plane and the data plane, enabling centralized, flexible, and programmable network control. Although local area networks like data center networks have benefited from SDN, it is still a problem to deploy SDN in wide area networks (WANs) or large-scale networks. Existing works show that multiple controllers are required in WANs with each covering one small SDN domain. However, the problems of SDN domain partition and controller placement should be further addressed. Therefore, we propose the spectral clustering based partition and placement algorithms, by which we can partition a large network into several small SDN domains efficiently and effectively. In our algorithms, the matrix perturbation theory and eigengap are used to discover the stability of SDN domains and decide the optimal number of SDN domains automatically. To evaluate our algorithms, we develop a new experimental framework with the Internet2 topology and other available WAN topologies. The results show the effectiveness of our algorithm for the SDN domain partition and controller placement problems.     

Throughput in Processor-Sharing Queues

Processor-sharing queues are often used to model file transmission in networks. While sojourn time is a common performance metric in the queueing literature, average transmission rate is the more commonly discussed metric in the networking literature. Whereas much is known about sojourn times, there is little known about the average service rate experienced by jobs in processor-sharing queues. We first define the average rate as observed by users and by the queue. In an M/M/1 processor-sharing queue, we give closed-form expressions for these average rates, and prove a strict ordering amongst them. We prove that the queue service rate (in bps) is an increasing function of the minimum required average transmission rate, and give a closed-form expression for the marginal cost associated with such a performance requirement. We then consider the effect of using connection access control by modeling an M/M/1/K processor-sharing queue. We give closed-form expressions for average transmission rates, and discuss the relationship between the queue service rate (in bps), the queue limit, the average rate, and the blocking probability