面向智慧城市的网络性能监控及流量预测研究

发布时间：2019-04-04 14:46

【摘要】：伴随着Intenet高速发展,网络应用不断完善,应用范围不断扩大,城市建设也迎来了智慧城市这个新的发展模式,网络设备散布于城市的每个角落,给人们的生活带来了很大的方便,网络设备的增多给网络管理带来了愈来愈大的压力。自从网络产生以来,网络管理系统已经迭代过多个版本,网络提供商推出的每一代网络管理系统适应了当时的需求,随着智慧城市进程的推进,迫切需要开发一套面向智慧城市的网络性能监控及流量预测系统,符合智慧城建设的现实需求,为智慧城市网络建设提供理论支撑,推动智慧城市建设不断向前发展。本论文根据实现面向智慧城市建设的网络性能监控及网络流量预测的工作需求,提出并实现了一种适用于智慧城市的网络性能监控及网络流量预测的仿真研究。本文网络性能监控系统的实现,以SNMP、IPMI等协议为数据采集模块基础,对采集的数据进行处理,方便前台页面生成告警,网络管理人员通过配置策略可在前台页面对网络数据采集条件进行配置,并提出了一种面向智慧城市的网络数据采集的格式化入库模型。详细阐述了网络性能监控模型的架构、每一个模块的功能及web形式作为监控系统的优势。本论文在网络流量的预测仿真方面,在详细了解遗传算法的特点和RBF神经网络的基本原理后,设计了一个利用改进的遗传算法来提高RBF神经网络的整体性能的思路,来建立一个网络流量预测模型,此模型以影响网络流量的因子为输入,网络流量为输出。本文设计实现的面向智慧城市的网络性能监控及流量预测系统,经过本文性能监控系统采集并且录入数据库中的历史网络流量数据的验证,可以达到预想的效果。最后,本文的研究成果可以作为网络运营商提供可视化的网络性能监控和网络流量预测的模型,为面向智慧城市建设的网络管理提供理论依据。
[Abstract]:With the rapid development of Intenet, the network application continues to improve, the scope of application continues to expand, urban construction also ushered in the smart city this new development mode, the network equipment is scattered in every corner of the city. It brings great convenience to people's life. The increase of network equipment brings more and more pressure to network management. Since the emergence of the network, the network management system has iterated over several versions, and each generation of network management system introduced by the network provider has adapted to the needs of the time. With the advancement of the intelligent city process, There is an urgent need to develop a set of intelligent city-oriented network performance monitoring and traffic prediction system, which accords with the practical needs of smart city construction, provides theoretical support for intelligent city network construction, and promotes the continuous development of intelligent city construction. According to the requirement of network performance monitoring and network traffic prediction for smart city construction, a simulation study on network performance monitoring and network traffic prediction for smart city is proposed and implemented in this paper. In this paper, the realization of network performance monitoring system, based on SNMP,IPMI and other protocols for data acquisition module, to process the collected data, convenient for foreground pages to generate alarms, Network managers can configure the network data acquisition conditions in front of the front page through the configuration strategy, and put forward a formatted storage model for intelligent city-oriented network data acquisition. The architecture of the network performance monitoring model, the function of each module and the advantages of web as the monitoring system are described in detail. In the aspect of prediction and simulation of network traffic, after knowing the characteristics of genetic algorithm and the basic principle of RBF neural network in detail, this paper designs a train of thought to improve the overall performance of RBF neural network by using the improved genetic algorithm. To establish a network traffic prediction model, this model takes the factors that affect the network traffic as input and the network traffic as the output. The intelligent city-oriented network performance monitoring and traffic prediction system is designed and implemented in this paper. After collecting and inputting the historical network traffic data in the database, the system can achieve the desired results. Finally, the research results of this paper can be used as a model for network operators to provide visual network performance monitoring and network traffic prediction, and provide a theoretical basis for network management oriented to intelligent city construction.
【学位授予单位】：沈阳理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.06

【参考文献】