基于黄金分割搜索算法的网络流量赫斯特指数计算与GUI系统设计

发布时间：2018-06-23 22:36

本文选题：长相关 + 网络流量　；参考：《华东师范大学》2017年硕士论文

【摘要】：网络流量一直是网络研究领域的重点之一,网络流量的研究对于了解网络的行为,提高网络的性能,以及保障网络的安全具有重要意义。网络流量具有自相似性,是典型的长相关信号,许多模型已经被应用于网络流量的研究,而且这些模型都有共同的核心参数,即赫斯特指数。赫斯特指数不仅对于网络流量建模具有十分重要的意义,而且对于研究网络流量的特性也具有很重要的参考价值。专家学者们已经提出了许多赫斯特指数的估计方法,但有些方法在计算效率方面存在一些局限,严重影响某些领域和场合下对于赫斯特指数计算的高时效性的需求。因此,具有较高计算效率和准确性的赫斯特指数估计算法对于网络流量的研究具有重要意义。本文第一章介绍了网络流量的赫斯特指数估计算法的研究意义以及国内外的研究现状;第二章介绍了传统的赫斯特指数估计方法并分析了其中一些方法在计算效率上存在的不足;第三章和第四章针提出了黄金分割搜索算法和随机搜索算法对传统算法进行改进;在第五章使用实际网络流量应用本文涉及到的算法进行实验并比较这些算法在计算效率上的差异;第六章在MATLAB的GUI平台上设计了一款网络流量赫斯特指数估计软件,为网络流量赫斯特指数的计算提供了一款方便快捷的科学计算工具。本文的主要贡献有:(1)针对传统方法的不足提出黄金分割搜索算法和随机搜索算法进行改进;(2)基于局部均值分解算法进行赫斯特指数估计;(3)在MATLAB的GUI平台上设计了一款网络流量赫斯特指数估计系统。
[Abstract]:Network traffic has always been one of the key points in the field of network research. The study of network traffic is of great significance to understand the behavior of the network, improve the performance of the network, and ensure the security of the network. Network traffic is self-similar, and it is a typical long correlation signal. Many models have been applied to the research of network traffic, and these models have a common core parameter, namely, Hurst index. Hurst exponent is not only of great significance for network traffic modeling, but also of great reference value for studying the characteristics of network traffic. Experts and scholars have put forward many estimation methods of Hurst exponent, but some methods have some limitations in computing efficiency, which seriously affect the demand for high time-efficiency of Hurst exponent calculation in some fields and situations. Therefore, the Hurst exponent estimation algorithm with high computational efficiency and accuracy is of great significance to the research of network traffic. The first chapter of this paper introduces the research significance of the Hurst exponent estimation algorithm of network traffic and the research status at home and abroad. The second chapter introduces the traditional Hurst index estimation method and analyzes the shortcomings of some of the methods in computing efficiency. Chapter 3 and chapter 4 propose golden section search algorithm and random search algorithm to improve the traditional algorithm. In the fifth chapter, we use the actual network traffic to use the algorithms mentioned in this paper to experiment and compare the computational efficiency of these algorithms. Chapter 6 designs a software for estimating the Hurst exponent of network traffic on the GUI platform of MATLAB. It provides a convenient and quick scientific calculation tool for calculating the Hurst index of network traffic. The main contributions of this paper are as follows: (1) the golden section search algorithm and random search algorithm are improved in view of the shortcomings of traditional methods; (2) the Hurst exponent estimation based on local mean decomposition algorithm; (3) designed on the GUI platform of MATLAB. A network traffic Hurst index estimation system.
【学位授予单位】：华东师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.06

【相似文献】