A sufficient condition and a necessary condition for the diffusion approximations of multiclass queueing networks under priority service disciplines

We establish a sufficient condition for the existence of the (conventional) diffusion approximation for multiclass queueing networks under priority service disciplines. The sufficient condition relates to a sufficient condition for the weak stability of the fluid networks that correspond to the queueing networks under consideration. In addition, we establish a necessary condition for the network to have a continuous diffusion limit; the necessary condition is to require a reflection matrix (of dimension equal to the number of stations) to be completely-S. When applied to some examples, including generalized Jackson networks, single station multiclass queues, first-buffer-first-served re-entrant lines, a two-station Dai–Wang network and a three-station Dumas network, the sufficient condition coincides with the necessary condition.     

Strong approximations for priority queues; head-of-the-line-first discipline

In this paper, we obtain strong approximation theorems for a single server queue withr priority classes of customers and a head-of-the-line-first discipline. By using priority queues of preemptive-resume discipline as modified systems, we prove strong approximation theorems for the number of customers of each priority in the system at timet, the number of customers of each priority that have departed in the interval [0,t], the work load in service time of each priority class facing the server at timet, and the accumulated time in [0,t] during which there are neither customers of a given priority class nor customers of priority higher than that in the system.Research supported by the National Natural Science Foundation of China.     

Diffusion approximations for re-entrant lines with a first-buffer-first-served priority discipline

The diffusion approximation is proved for a class of queueing networks, known as re-entrant lines, under a first-buffer-first-served (FBFS) service discipline. The diffusion limit for the workload process is a semi-martingale reflecting Brownian motion on a nonnegative orthant. This approximation has recently been used by Dai, Yeh and Zhou [21] in estimating the performance measures of the re-entrant lines with a FBFS discipline.Supported in part by a grant from NSERC (Canada).Supported in part by a grant from NSERC (Canada); the research was done while the author was visiting the Faculty of Commerce and Business Administration, UBC, Canada.     

具有优先权的Lu-Kumar排队网络的扩散逼近

对Lu-Kumar排队网络来说,标准的额定负荷条件,即每个工作站的工作强度ρ<1,并不足以保证该排队网络的稳定性,特别是在具有优先权的服务规则下.论文在讨论了Lu-Kumar排队网络稳定性相关结果的基础上,研究了Lu-Kumar排队网络在具有优先权的服务规则下的扩散逼近.证明了当每个工作站的额定负荷ρ→1时,Lu-Kumar排队网络对具有优先权的服务规则的所有优先级别来说,扩散逼近定理均成立.     

一类重入型网络在优先服务原则下的扩散近似

本文研究了一类重入型网络在优先服务原则下的扩散近似,运用随机分析方法,证明了标准化队长过程的C-紧性．在优先服务原则下,给出了这类网络的标准化队长过程扩散近似存在的充分条件．     

On the fluid approximation for a multiclass queue under non-preemptive SBP service discipline

A multi-class single server queue under non-preemptive static buffer priority (SBP) service discipline is considered in this paper. Using a bounding technique, we obtain the fluid approximation for the queue length and busy time processes. Furthermore, we prove that the convergence rate of the fluid approximation for the queue length and busy time processes is exponential for large N. Additionally, a sufficient condition for stability is obtained.     

Strong approximations for Markovian service networks

Inspired by service systems such as telephone call centers, we develop limit theorems for a large class of stochastic service network models. They are a special family of nonstationary Markov processes where parameters like arrival and service rates, routing topologies for the network, and the number of servers at a given node are all functions of time as well as the current state of the system. Included in our modeling framework are networks of M _t /M _t /n _t queues with abandonment and retrials. The asymptotic limiting regime that we explore for these networks has a natural interpretation of scaling up the number of servers in response to a similar scaling up of the arrival rate for the customers. The individual service rates, however, are not scaled. We employ the theory of strong approximations to obtain functional strong laws of large numbers and functional central limit theorems for these networks. This gives us a tractable set of network fluid and diffusion approximations. A common theme for service network models with features like many servers, priorities, or abandonment is “non-smooth” state dependence that has not been covered systematically by previous work. We prove our central limit theorems in the presence of this non-smoothness by using a new notion of derivative. This revised version was published online in June 2006 with corrections to the Cover Date.     

Strong approximations and sequential change-point analysis for diffusion processes

In this paper ergodic diffusion processes depending on a parameter in the drift are considered under the assumption that the processes can be observed continuously. Strong approximations by Wiener processes for a stochastic integral and for the estimator process constructed by the one-step procedure of Le Cam are obtained. Applying these approximations, a CUSUM-type procedure is developed for the sequential testing of changes in the parameter.     

The throughput performance of a prioritized LIFO service discipline

The LIFO service discipline maximizes throughput for certain queuing systems with delay-dependent customer behavior. To study the effects of priorities in such a system, we obtain delay distributions for each customer class of a K-class, single server system with nonpreemptive priorities and LIFO within each queue.     

非强占FBFS服务规则下Re-entrant Line排队网络的扩散逼近

本文研究了-个非强占静态优先权first-buffer-first-served(FBFS)服务规则下的re-entrant line排队网络.文章首先建立了-个极限定理,后通过分析队长和斜反射映射的关系,建立了队长过程和闲期过程的扩散逼近.     

Existence Condition for the Diffusion Approximations of Multiclass Priority Queueing Networks

In this paper, we extend the work of Chen and Zhang [12] and establish a new sufficient condition for the existence of the (conventional) diffusion approximation for multiclass queueing networks under priority service disciplines. This sufficient condition relates to the weak stability of the fluid networks and the stability of the high priority classes of the fluid networks that correspond to the queueing networks under consideration. Using this sufficient condition, we prove the existence of the diffusion approximation for the last-buffer-first-served reentrant lines. We also study a three-station network example, and observe that the diffusion approximation may not exist, even if the proposed limiting semimartingale reflected Brownian motion (SRBM) exists.     

Queues with group arrivals and exhaustive service discipline

Queues with compound Poisson arrivals, phase-type service and exhaustive service discipline are studied. An algorithmic method is developed to compute the steady-state probability distribution of the number of customers in the system with unlimited or limited queue capacities. Examples with different model parameters are given to show the computational efficiency of the method. In the Appendix, the stochastic decomposition property for the queues with single arrivals and with exhaustive service discipline is extended to queues with group arrivals.     

离散指数分布序列的强偏差与强逼近

本文引入离散指数分布概念,建立了关于离散型指数分布序列的强偏差定理和强大数定律.同时,得到离散指数分布序列对连续指数分布序列的强逼近.     

A large closed queueing network with autonomous service and bottleneck

This paper studies the queue-length process in a closed Jackson-type queueing network with the large number N of homogeneous customers by methods of the theory of martingales and by the up- and down-crossing method. The network considered here consists of a central node (hub), being an infinite-server queueing system with exponentially distributed service times, and k single-server satellite stations (nodes) with generally distributed service times with rates depending on the value N. The service mechanism of these k satellite stations is autonomous, i.e., every satellite server j serves the customers only at random instants that form a strictly stationary and ergodic sequence of random variables. Assuming that the first k-1 satellite stations operate in light usage regime the paper considers the cases where the kth satellite station is a bottleneck node. The approach of the paper is based both on development of the method from the paper by Kogan and Liptser [16], where a Markovian version of this model has been studied, and on development of the up- and down-crossing method. This revised version was published online in June 2006 with corrections to the Cover Date.     

高负荷下带重尾服务强占优先排队的扩散逼近

考虑的排队系统是单服务台,顾客的初始到来是依泊松过程来到服务台,顾客的服务时间是重尾分布,服务的原则是强占优先服务.在高负荷条件下对此模型进行研究,获得了系统中的负荷过程,离去过程和队长过程的扩散逼近.     

Analysis of a time-limited service priority queueing system with exponential timer and server vacations

We consider a multi-class priority queueing system with a non-preemptive time-limited service controlled by an exponential timer and multiple (or single) vacations. By reducing the service discipline to the Bernoulli schedule, we obtain an expression for the Laplace-Stieltjes transform (LST) of the waiting time distribution via an iteration procedure, and a recursive scheme to calculate the first two moments. It is noted that we have to select embedded Markov points based on the service beginning epochs instead of the service completion epochs adopted for most of M/G/1 queueing analyses. Through the queue-length analysis, we obtain a decomposition form for the LST of the waiting time in each queue having the exhaustive service.      

Analysis of a discrete-time preemptive resume priority buffer

In this paper, we analyze a discrete-time preemptive resume priority queue. We consider two classes of customers which have to be served, where customers of one class have preemptive resume priority over customers of the other. Both classes contain customers with generally distributed service times. We show that the use of probability generating functions is beneficial for analyzing the system contents and customer delays of both classes. It is shown (theoretically as well as by some practical procedures) how moments and approximate tail probabilities of system contents and customer delays are calculated. The influence of the priority scheduling discipline and the service time distributions on the performance measures is shown by some numerical examples.     

M/G/1/N vacation model with varying E-limited service discipline

We define and analyze anM/G/1/N vacation model that uses a service discipline that we call theE-limited with limit variation discipline. According to this discipline, the server provides service until either the system is emptied (i.e. exhausted) or a randomly chosen limit ofl customers has been served. The server then goes on a vacation before returning to service the queue again. The queue length distribution and the Laplace-Stieltjes transforms of the waiting time, busy period and cycle time distributions are found. Further, an expression for the mean waiting time is developed. Several previously analyzed service disciplines, including Bernoulli scheduling, nonexhaustive service and limited service, are special cases of the general varying limit discipline that is analyzed in this paper.     

A Strong Approximation Theorem for Quasi-associated Sequences

By combining the Csorgo-Révész quantile transform methods and the Skorohod-Strassen martingale embedding theorem, we prove a strong approximation theorem for quasi-associated random variables with mean zero and finite （2 ＋ δ）th moment under polynomial decay rate. As a consequence, the decay rate for a strong approximation theorem of associated sequences of Yu （1996） is weakened.     

Fluid approximations for a processor-sharing queue

In this paper a fluid approximation, also known as a functional strong law of large numbers (FSLLN) for a GI/G/1 queue under a processor-sharing service discipline is established and its properties are analysed. The fluid limit depends on the arrival rate, the service time distribution of the initial customers, and the service time distribution of the arriving customers. This is in contrast to the known result for the GI/G/1 queue under a FIFO service discipline, where the fluid limit is piecewise linear and depends on the service time distribution only through its mean. The piecewise linear form of the limit can be recovered by an equilibrium type choice of the initial service distribution. This revised version was published online in June 2006 with corrections to the Cover Date.