带有不耐烦顾客和非零匹配时间的双边排队系统的尾渐近分析

Exact Tail Asymptotics for a Double-ended Queue with Nonpersistent Customers and Nonzero Matching Time

  • 摘要: 我们考虑一类带有不耐烦顾客和非零匹配时间的双边排队系统,研究其处于平稳状态时各自队长的边界概率、边缘概率以及联合概率的尾部渐近性质.本文的研究思路是将排队过程看作四分之一平面内的随机游动,利用核方法,我们详细讨论了未知生成函数的控制奇点的位置并确定了生成函数在控制奇点处的极限性质,最后运用Tauberian-Like定理,得到相关平稳分布的尾渐近性质.

     

    Abstract: In this paper, we study a double-ended queueing system with nonzero matching time and nonpersistent customers. For this model, we aim at studying the exact tail asymptotics for the boundary distribution, the marginal distribution and the joint distribution of the queue length, respectively. We model the queueing process as a random walk in the quarter plane. By applying the kernel method, we firstly determine the location of the dominated singularity of the unknown generating function. Then we analyze the asymptotic behaviors of the generating function at the dominated singularity. Finally, we obtain the exact tail asymptotics of the stationary distributions by using Tauberian-like theorem.

     

/

返回文章
返回