距离矢量路由混合SDN网络的一致性更新

混合SDN网络作为SDN网络与传统网络的过渡形态,在保留传统网络服务的同时,兼有部分软件定义网络易于管理维护的特点,但混合SDN网络的更新问题一直没能找到有效的解决方法,特别是当距离矢量路由参与混合的情况.本文将分段路由技术引入混合SDN网络更新,提出了一种分段路由在混合SDN网络更新的应用方案,结合顺序更新与两步复制机制,设计LFCA算法.LFCA算法首先尝试用分段路由拼接最终路径并将最终路径信息封装入数据包包头.而对于不可拼接的数据流,LFCA算法构建了顾及更新远端效应的动态拓扑集并基于回溯原理来评估每一步更新操作,通过不断回溯比较来最终找到最长一致性更新序列,对于仍未完成更新的节点利用两步复制机制完成更新.经过实验验证,与之前算法相比,LFCA算法可以极大地节省算法时间开销与存储资源开销.

混合SDN的多目标路由优化算法

为充分利用SDN(Software Defined Network,软件定义网络)节点的优势来优化网络性能,混合SDN的流量工程成为当前的研究热点,而路由优化是实现流量工程目标的重要策略之一。但是,当前混合SDN的流量工程中未考虑网络整体的负载均衡以及SDN节点的处理能力。针对上述问题,提出了一种多目标路由优化算法——MCS(Minimum Cost Sum,最小化代价和),综合考虑整个网络的传输延迟与链路利用率,同时保证SDN节点的处理能力满足实际约束,最终实现全网综合性能的最优化。实验结果表明,当网络整体负载较轻,MCS与现有的SOTE(SDN/OSPF Traffic Engineering,软件定义网络/开放最短路径优先流量工程)算法性能相近;而当网络负载加重时,MCS相比于SOTE,可将网络负载降低约9%,因此,MCS算法具有更高的优化能力。

A SDN-Based Energy Saving Strategy in Wireless Access Networks

This paper investigates on the base stations(BSs) sleeping control and energy saving in wireless network. The objective is to find the sleeping control and energy saving configuration between total power consumption and average video's quality. On the Software Defined Network(SDN) access network architecture, a type of sleeping control and active BSs' optimal transmitting time strategy is considered, the BS sleeps when there is no active users, and wakes up after a period of vacation time. In this paper, we study the active users grouping strategy, In order to spare more BSs into sleeping mode. Then this paper proposes an active BS transmitting time optimal strategy according to the users' Qo S. In the proposed strategy, the active BSs' transmitting time is minimized in order to save energy. This paper employs the mixed integer-programming model to present this optimization problem. Then we utilized a novel algorithm to save the energy in access networks and also meet the Qo S requirements. Both the analytical and simulation results show that the algorithm can effectively save energy in the access network BSs.