Convergence rate for the distributions of GI/M/1/n and M/GI/1/n as n tends to infinity

被引:8
作者
Simonot, F [1 ]
机构
[1] ESSTIN, F-54500 Vandoeuvre Nancy, France
关键词
GI/M/1/n; M/GI/1/n; Markov chain; Lindley process; coupling; convergence rate;
D O I
10.2307/3215017
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
In this note, we compare the arrival and time stationary distributions of the number of customers in the GI/M/1/n and GI/M/1 queueing systems. We show that, if the interarrival c.d.f. H is non-lattice with mean value lambda(-1), and if the traffic intensity rho=lambda mu(-1) is strictly less than one, then the convergence rate of the stationary distributions of GI/M/1/n to the corresponding stationary distributions of GI/M/1 is geometric. Moreover, the convergence rate can be characterized by the number omega, the unique solution in (0, 1) of the equation z=integral(0)(infinity)exp{-mu(1-z)t}dH(t). A similar result is established for the M/GI/1/n and M/GI/1 queueing systems.
引用
收藏
页码:1049 / 1060
页数:12
相关论文
共 14 条
[1]   EQUILIBRIUM PROPERTIES OF THE M-G-1 QUEUE [J].
ASMUSSEN, S .
ZEITSCHRIFT FUR WAHRSCHEINLICHKEITSTHEORIE UND VERWANDTE GEBIETE, 1981, 58 (02) :267-281
[2]  
Asmussen S, 2008, APPL PROBABILITY QUE, V51
[3]  
Billingsley P, 1968, CONVERGE PROBAB MEAS
[4]   STOCHASTICALLY MONOTONE MARKOV CHAINS [J].
DALEY, DJ .
ZEITSCHRIFT FUR WAHRSCHEINLICHKEITSTHEORIE UND VERWANDTE GEBIETE, 1968, 10 (04) :305-&
[5]   MONOTONE INFINITE STOCHASTIC MATRICES AND THEIR AUGMENTED TRUNCATIONS [J].
GIBSON, D ;
SENETA, E .
STOCHASTIC PROCESSES AND THEIR APPLICATIONS, 1987, 24 (02) :287-292
[6]  
Heyman D. P., 1989, Queueing Systems Theory and Applications, V5, P381, DOI 10.1007/BF01225326
[8]  
KALASHNIKOV VV, 1990, MATH METHODS CONSTRU
[9]  
Rachev S.T., 1991, PROBABILITY METRICS