Retrial queue

About: Retrial queue is a research topic. Over the lifetime, 784 publications have been published within this topic receiving 12354 citations.

Two way communication retrial queues with balanced call blending

Abstract: In call centers, call blending consists in the mixing of incoming and outgoing call activity. Artalejo and Phung-Duc recently provided an apt model for such a setting, with a two way communication retrial queue. However, by assuming a classical (proportional) retrial rate for the incoming calls, the outgoing call activity is largely blocked when many incoming calls are in orbit, which may be unwanted, especially when outgoing calls are vital to the service offered. In this paper, we assume a balanced way of call blending, through a retrial queue with constant retrial rate for incoming calls. For the single server case (one operator), a generating functions approach enables deriving explicit formulas for the joint stationary distribution of the number of incoming calls and the system state, and also for the factorial moments. This is complemented with a stability analysis, expressions for performance measures, and also recursive formulas, allowing reliable numerical calculation. For the multiserver case (multiple operators), we provide a quasi-birth-and-death process formulation, enabling deriving a sufficient and necessary condition for stability in this case, as well as a numerical recipe to obtain the stationary distribution.

Some Features of a Finite-Source M/GI/1 Retrial Queuing System with Collisions of Customers

Abstract: In this paper a finite-source M/GI/1 retrial queuing system with collisions of customers is considered. The definition of throughput of the system as average number of customers, which are successfully served per unit time is introduced. It is shown that at some combinations of system parameter values and probability distribution of service time of customers the throughput can be arbitrarily small, and at another values of parameters throughput can be greater than the service intensity. Applying method of asymptotic analysis under the condition of unlimited growing number of sources it is proofed that limiting distribution of the number of retrials/transitions of the customer into the orbit is geometric and the sojourn/waiting time of the customer in the orbit follows a generalized exponential distribution. In addition, the mean sojourn time of the customer under service is obtained.

Steady state analysis of batch arrival feedback retrial queue with two phases of service, negative customers, Bernoulli vacation and server breakdown

Abstract: In steady state, a batch arrival feedback retrial queue with negative customers has been discussed. Any arriving batch of positive customers finds the server is free, one of the customers from the batch enters into the service area and the rest of them join into the orbit/retrial group, where the server provides two essential phases of service to each positive customer. The negative customer, arriving during the service time of a positive customer, will remove the positive customer in-service and the interrupted positive customer either enters into the orbit with probability θ or leaves the system with probability 1-θ. The busy server may breakdown at any instant and the service channel will fail for a short interval of time. The steady state probability generating function for the system size is obtained by using the supplementary variable method. Numerical illustrations are analysed to see the effect of system parameters.

Tool supported modeling of sensor communication networks by using finite-source priority retrial queues

Abstract: The main aim of the present paper is to draw the attention of the readers of this special issue to the modeling issues of sensor networks. The novelty of this investigation is the introduction of servers vacation combined with priority customers for finite-source retrial queues and its application to wireless sensor networks. In this paper we analyze a priority finite-source retrial queue with repeated vacations. Two types of priority customers are defined, customers with priority 1 (P1) go directly to an ordinary FIFO queue. However, if customers with priority 2 (P2) find the server in busy or unavailable state go to the orbit. These customers stay in the orbit and retry their request until find the server in idle and available state. We assume that P1 customers have non-preemptive priority over P2 customers. The server starts with a listening period and if no customer arrive during this period it will enter in the vacation mode. When the vacation period is terminated, then the node wakes up. If there is a P1 customer in the queue the server begin to serve it, and when there is no any P1 customer, the node will remain awake for exponentially distributed time period. If that period expires without arrivals the node will enter in the next sleeping period. All random variables involved in model construction are supposed to be independent and exponentially distributed ones. Our main interest is to give the main steady-state performance measures of the system computed by the help of the MOSEL tool. Sev∗Acknowledgment: This research is partially supported by the Hungarian Science and Technology Foundation, HungarianFrench Bilateral Cooperation under grant TeT 10-1-2011-0741, FR25/2010. eral Figures illustrate the effect of input parameters on the mean response time.