基于智能SDN的无线网络控制机制研究
发布时间：2018-12-17 19:22
【摘要】：软件定义网络(Software-Defined Networking)是一种新型的网络架构,通过控制面和数据面分离、集中式的管理和可编程的接口实现对网络的灵活控制。将SDN和无线网络结合受到越来越多的关注,本文针对基于SDN的无线网络控制机制展开研究。在对已有研究成果对比和分析的基础上,提出了基于智能SDN的无线局域网管理方案,通过对OpenFlow协议的扩展实现SDN控制器对无线AP和交换机的统一管理,解决大规模无线局域网组网面临的集中式接入管理、移动性管理和干扰管理等问题,同时将智能中心引入SDN,解决了控制器功能不足以及控制器之间协同的问题。针对切换问题,提出了三层无缝切换机制,通过Mininet仿真验证方案在减少切换时延上的有效性。针对干扰问题提出了基于干扰图和流量预测的信道分配算法,仿真结果表明算法能够有效减少系统中无线AP之间的干扰,提升系统的性能。针对蜂窝网络向异构、融合、多租户承载等方向演进的需求,提出了一种基于智能SDN的异构蜂窝网络管理架构,将蜂窝网中的无线接入网和核心网进行统一管理和协同,基于智能决策中心,实现无线接入网资源的动态适配。针对多租户业务流转发带来的路径分配问题,提出了基于增强学习的网络切片算法和路径分配算法,在保证租户间隔离的同时提供对业务的QoS保障,通过仿真验证了算法的有效性。
[Abstract]:Software defined Network (Software-Defined Networking) is a new network architecture, which can control the network flexibly through the separation of control surface and data surface, centralized management and programmable interface. More and more attention is paid to the combination of SDN and wireless network. In this paper, the control mechanism of wireless network based on SDN is studied. Based on the comparison and analysis of the existing research results, a WLAN management scheme based on intelligent SDN is put forward. Through the extension of OpenFlow protocol, the unified management of wireless AP and switch by SDN controller is realized. In order to solve the problems of centralized access management, mobility management and interference management in large-scale WLAN network, the intelligent center is introduced into SDN, to solve the problems of insufficient function of controller and coordination between controllers. Aiming at the handoff problem, a three-layer seamless handoff mechanism is proposed, and the effectiveness of the scheme in reducing handoff delay is verified by Mininet simulation. A channel assignment algorithm based on interference graph and traffic prediction is proposed. The simulation results show that the algorithm can effectively reduce the interference between wireless AP in the system and improve the performance of the system. Aiming at the evolution of cellular network towards heterogeneity, convergence and multi-tenant bearing, a heterogeneous cellular network management architecture based on intelligent SDN is proposed, which manages and collaborates the wireless access network and core network in cellular network. Based on intelligent decision center, the dynamic adaptation of wireless access network resources is realized. In order to solve the path assignment problem caused by multi-tenant traffic forwarding, a network slicing algorithm and a path assignment algorithm based on reinforcement learning are proposed, which can guarantee the separation between tenants and provide the QoS guarantee for the service at the same time. The effectiveness of the algorithm is verified by simulation.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN92
本文编号：2384686
[Abstract]:Software defined Network (Software-Defined Networking) is a new network architecture, which can control the network flexibly through the separation of control surface and data surface, centralized management and programmable interface. More and more attention is paid to the combination of SDN and wireless network. In this paper, the control mechanism of wireless network based on SDN is studied. Based on the comparison and analysis of the existing research results, a WLAN management scheme based on intelligent SDN is put forward. Through the extension of OpenFlow protocol, the unified management of wireless AP and switch by SDN controller is realized. In order to solve the problems of centralized access management, mobility management and interference management in large-scale WLAN network, the intelligent center is introduced into SDN, to solve the problems of insufficient function of controller and coordination between controllers. Aiming at the handoff problem, a three-layer seamless handoff mechanism is proposed, and the effectiveness of the scheme in reducing handoff delay is verified by Mininet simulation. A channel assignment algorithm based on interference graph and traffic prediction is proposed. The simulation results show that the algorithm can effectively reduce the interference between wireless AP in the system and improve the performance of the system. Aiming at the evolution of cellular network towards heterogeneity, convergence and multi-tenant bearing, a heterogeneous cellular network management architecture based on intelligent SDN is proposed, which manages and collaborates the wireless access network and core network in cellular network. Based on intelligent decision center, the dynamic adaptation of wireless access network resources is realized. In order to solve the path assignment problem caused by multi-tenant traffic forwarding, a network slicing algorithm and a path assignment algorithm based on reinforcement learning are proposed, which can guarantee the separation between tenants and provide the QoS guarantee for the service at the same time. The effectiveness of the algorithm is verified by simulation.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN92
【参考文献】
相关期刊论文 前1条
1 周逸凡;赵志峰;张宏纲;;基于智能SDN面向5G的异构蜂窝网络架构[J];电信科学;2016年06期
，本文编号：2384686
本文链接：https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2384686.html
