CAJ | 학술논문

该文基于混沌理论提出了一种使用海量网络流量数据对大规模网络性能进行有效评估的方法.在长期链路利用率数据呈现出明显的周期性行为,和短期链路利用率数据具有混沌特征的前提下,选取最大Lyapunov指数作为一项性能评估参数来评估网络性能.分析结果表明最大Lyapunov指数较常见统计量如数学期望、方差等更能有效反映流量的行为趋势.
해문기우혼돈이론제출료일충사용해량망락류량수거대대규모망락성능진행유효평고적방법.재장기련로이용솔수거정현출명현적주기성행위,화단기련로이용솔수거구유혼돈특정적전제하,선취최대Lyapunov지수작위일항성능평고삼수래평고망락성능.분석결과표명최대Lyapunov지수교상견통계량여수학기망、방차등경능유효반영류량적행위추세.
A method based on chaos theory is presented in this paper, to evaluate large-scale network performance using massive traffic measurement. As the periodicity of long-term link utilization measurement and the chaotic nature of short-term link utilization measurement, the largest Lyapunov exponent can be selected as a performance evaluation parameter to represent the performance of network. Analysis results show that the largest Lyapunov exponent can achieves better results than commonly used statistics, such as mathematical expectation and variance.