论文部分内容阅读
研究了高速网络中的流量矩阵估计问题.针对该问题在时域中的高病态特性,提出一种新的算法来解决该问题,该算法利用广义S变换和压缩感知来建模流量矩阵.通过广义S变换,则在时频域中对流量矩阵进行估计;在时频域中流量矩阵被分解为稳态分量和波动分量两部分,稳态分量通过抽样测量的样本数据进行建模分析,而波动分量则利用压缩感知理论进行估计.最后,利用真实网络流量数据验证所提出的算法,仿真结果表明所提出的算法是有效和可行的.
In this paper, we study the problem of traffic matrix estimation in high-speed networks.A new algorithm is proposed to solve this problem in the time-domain with high ill-posedness.The algorithm uses generalized S-transform and compressed sensing to model the traffic matrix, The generalized S transform estimates the traffic matrix in the time-frequency domain. The traffic matrix is decomposed into the steady-state component and the fluctuating component in the time-frequency domain. The steady-state component is modeled by the sampled sample data Volatile components are estimated using the compressed sensing theory.Finally, the proposed algorithm is verified by real network traffic data, and the simulation results show that the proposed algorithm is effective and feasible.