论文部分内容阅读
为了刻画网络冗余流量在小时间尺度下的局部特性,提出了基于柯西-拉普拉斯多分形小波模型及其参数估计算法.采用联合分布函数来描述冗余流量的局部表征,即分别用柯西分布和拉普拉斯分布计算冗余流量的重尾和尖峰参数乘法因子;通过概率比较方法获取小波系数和尺度系数的比例参数阈值,以界定两种不同分布的参数范围.实验结果表明,提出的模型能准确且有效地描述网络冗余流量在小时间尺度下的多分形特性.
In order to characterize the local characteristics of network redundant traffic on a small time scale, a Cauchy-Laplacian multifractal wavelet model and its parameter estimation algorithm are proposed. The joint distribution function is used to describe the local characterization of redundant traffic, The Cauchy distribution and Laplace distribution were used to calculate the multiplicative factors of heavy-tailed and spiking parameters of redundant traffic, and the thresholds of proportion parameters of wavelet coefficient and scale coefficient were obtained by probability comparison method to define the parameter range of two different distributions.The experimental results It shows that the proposed model can accurately and effectively describe the multifractal characteristics of network redundant traffic on a small time scale.