基于SDN的可生存性增强自配置技术研究

发布时间：2018-03-13 04:28

本文选题：软件定义网络　切入点：认知网络　出处：《哈尔滨理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：近年来,信息技术高速发展,Internet已广泛应用于人们的日常生活。另外各类任务关键系统也逐渐趋于网络化,但是因为网络的开放性和复杂性、运行环境的多变性、系统的设计缺陷以及人为误操作等因素不可避免,导致任务关键系统面临巨大的风险,一旦失效,将会给国民经济和人民生活带来巨大的影响。然而人们依赖软件系统的程度却与日俱增,这就产生了人们高可信赖的需求和当前系统的低可靠性二者之间的矛盾。人们希望软件系统可以持续稳定地提供关键服务,即使是遭到了攻击或者是系统故障等情况,从而提高系统的可生存性。本文主要以增强系统的可生存性作为研究目标,首先在基于软件定义网络(Software Defined Networking,SDN)这一新型网络体系架构上,提出了一种基于蚁群混合遗传算法的虚拟网络映射算法,该算法将网络中的实体资源规约为运行情况下的可操纵资源;其次,考虑到一旦底层基础设施发生改变会导致原有可行方案失效的情况,改良出一种面向节点失效的虚拟网络重映射算法,该算法通过引入脱落因子这一指标对当前环境下的资源进行评估和分析,并将失效的虚拟节点和链路从原有的虚拟网络请求中分离开来,然后通过贪婪算法重新对其进行合理、高效地配置,从而实现虚拟网络请求;最后借鉴认知网络中的自配置思想,提出了一种资源自配置方法,该方法利用其感知到的系统内、外界环境和自身状态,对资源进行动态地配置,从而提高软件系统智能地适应环境的能力,最终实现其可生存性的增强。
[Abstract]:In recent years, with the rapid development of information technology, Internet has been widely used in people's daily life. In addition, all kinds of mission-critical systems are becoming more and more networked, but because of the openness and complexity of the network, the operating environment is changeable. The design defects and human error of the system inevitably lead to the mission-critical system facing a huge risk, once the failure, Will have a huge impact on the national economy and people's lives. But people are increasingly dependent on software systems. This creates a contradiction between the need for high reliability and the low reliability of the current system. It is hoped that software systems can provide critical services on a sustainable and stable basis, even in the event of attacks or system failures. In order to improve the survivability of the system, this paper mainly aims at enhancing the survivability of the system. Firstly, on the new network architecture based on software definition, Software Defined networking (SDN), In this paper, a virtual network mapping algorithm based on ant colony hybrid genetic algorithm is proposed. Considering that once the underlying infrastructure changes will lead to the failure of the original feasible scheme, an improved virtual network remapping algorithm for node failure is proposed. The algorithm evaluates and analyzes the resources in the current environment by introducing the shedding factor, and separates the invalid virtual nodes and links from the original virtual network requests. Finally, a resource self-configuration method is proposed, which makes use of its perceived system, external environment and its own state, and realizes the virtual network request efficiently, and uses the self-configuration idea of cognitive network for reference, and proposes a resource self-configuration method, which makes use of its perceived system, external environment and its own state. Resources are dynamically configured to improve the ability of software system to adapt to the environment intelligently and finally to enhance the survivability of the software system.
【学位授予单位】：哈尔滨理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TP393.0

【相似文献】