基于伪似然估计算法的网络延迟计算仿真

发布时间：2018-06-14 18:43

本文选题：伪似然估计 + 网络延迟　；参考：《计算机仿真》2013年10期

【摘要】：在网络实际应用过程中,由于网络规模不断扩大,网络复杂性不断提高,在网络数据传输过程中,需要获取网络中的节点性能参数。但是,网络中的节点存在较大的差异性,将造成节点的性能参数过于复杂,在通信过程中协调性较差,存在冲突现象,导致对这种延迟的估计出现偏差。提出了一种伪似然的网络延迟估计算法。利用小波域数据线性变换模型,计算网络状态参数,为网络延迟计算提供准确的数据基础。统计网络延迟的概率分布情况,将网络延迟计算问题,转化为伪似然估计问题,并对该问题进行求解处理,消除干扰的影响。实验结果表明,利用该算法进行网络延迟计算,能够在网络规模比较大、网络复杂性比较高的情况下有效地获取网络延迟程度,获取的结果与实际的网络延迟分布情况非常接近,真实的反映了网络延迟情况。
[Abstract]:In the process of network application, because of the continuous expansion of network scale and the increasing of network complexity, it is necessary to obtain the performance parameters of nodes in the network data transmission process. However, there are large differences in the nodes in the network, which will cause the performance parameters of the nodes to be too complex, the coordination in the communication process is poor, and there is a conflict phenomenon, which leads to the deviation of the estimation of this delay. A pseudo-likelihood network delay estimation algorithm is proposed. The data linear transformation model in wavelet domain is used to calculate the network state parameters, which provides an accurate data basis for network delay calculation. According to the probability distribution of network delay, the computing problem of network delay is transformed into a pseudo-likelihood estimation problem, and the problem is solved to eliminate the influence of interference. The experimental results show that the network delay degree can be obtained effectively when the network scale is large and the network complexity is high by using the algorithm. The obtained results are very close to the actual network delay distribution and reflect the network delay.
【作者单位】：华东交通大学现代教育技术中心;
【分类号】：TP393.0

【参考文献】