Heavy traffic analysis of a random walk on a lattice semi-strip
Heavy traffic analysis of a random walk on a lattice semi-strip
复制标题
格子半带上随机游走的重交通分析
DOI：
--
复制
发表时间：
1995
期刊：
影响因子：
--
通讯作者：
Gennadi Falin
中科院分区：
文献类型：
--
作者：
Gennadi Falin
我们考虑状态空间为非负整数和有限集的乘积的马尔可夫链。转移概率满足相对于第一个坐标的有限空间同质性的某些条件。当链是“几乎非遍历链”时，我们研究了不变测度的渐近行为。主要结果表明：在此条件下：(A)链的第一个尺度分量与第二个分量渐近独立；(B)尺度第一个分量是渐近指数的。这个极限分布的参数是根据所谓的诱导链的特征给出的(它描述了的随机动力学和漂移的前两个矩
We consider a Markov chain whose state space is a product of non-negative integers and a finite set. Transition probabilities satisfy certain conditions of a limited spacial homogeneity with respect to the first coordinate, . We investigate asymptotic behaviour of the invariant measure when the chain is “almost nonergodic”. The main result states that under this condition: (a) the scaled first component of the chain , and the second component , are asymptotically independent; (b) the scaled first component , is asymptotically exponential. The parameter of this limit distribution is given in terms of characteristics of the so-called induced chain (which describes the stochastic dynamics of and the first two moments of drift of