摘要
以更新理论和拉普拉斯-司阶梯变换为数学工具推导出各种on/off周期持续时间尾分布情况下,on-off源的长程相关结构.证明了on-off源的长程相关结构主要取决于尾分布最重的那个周期的尾分布特性,并通过构造随机on-off样本序列对该结论的正确性进行了验证.
By using the mathematical tools of renewal theory and Laplace-Stieltjes transform, the large-timescale covariance structure of the on-off sources with general distributed on/off duration was derived. It was proved that the large-timescale autocovariance structure of on-off sources is mainly determined by the duration whose tail is heavier. This conclusion is identified by calculating the autocovariance functions of the empirically constructed on-off sequences.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2004年第z1期83-86,共4页
Journal of Shanghai Jiaotong University
关键词
on-off源
重尾分布
自协方差
on-off sources
heavy-tailed distribution
autocovariance