Non-exponential Service Times Research Articles

Production Systems with Parallel Heterogeneous Servers of Limited Capacity: Accurate Modeling and Performance Analysis

Heterogeneous systems of limited capacity have general applications in manufacturing, but also in logistic or service systems due to the differences in server or workstation performance or work assignment; this is in close relationship with system flexibility, where saturation and blocking are ordinary situations of systems with high demand and limited capacity, and thus, accurate loss quantification is essential for performance evaluation. Multi-class systems of limited capacity have been studied much less than parallel homogeneous systems (Erlang models). In this context, accurate models for parallel heterogeneous ordered-entry systems were developed: without any prior queue, i.e., M/Mi/c/c, and with a k-capacity queue, i.e., M/Mi/c/c + k. These new matrix models gave an exact state formulation, and their accuracy was verified using discrete event simulation and comparison with literature results. Also, the effect of the queue capacity was studied in relationship to the pattern of service rates. Next, the heterogeneous recirculating system model was also developed with good approximation results. Finally, the proposed models were applied to evaluate systems with non-exponential service times using a new hybrid methodology by combining the Markovian model and the Monte Carlo method (MCM) for normal or lognormal service times, which also yielded useful good approximations to the simulated system.

Estimating the response time of a data-intensive computing environment

Introduction: The amount of digital data is constantly growing as well as the need for its storage and processing for various purposes. To conduct data analysis, high-performance computing environments associated with parallelization methods, and, accordingly, data-intensive applications are used. The lack of quality tools for evaluating the effectiveness of the process of parallel data processing or tasks leads to excessive allocation of resources. Purpose: To develop mathematical models of data-intensive computing environments and methods for their performance analysis, i.e., for estimating the average system response time based on the data on system performance at the level of subtask solving. Results: We present a mathematical model of a parallel computing system in the form of a queueing system with parallel query processing on various architectures, including non-Poisson input flow and non-exponential service times. As a method for analyzing the average response time, we use a combination of simulation modeling with one of the machine learning methods (artificial neural networks). The effectiveness of the method is confirmed by numerical experiments and depends neither on the type of input flow, nor on the type of distribution of query service times, nor on the number of servers in the nodes of the system. The approximation error of the average response time does not exceed 10%, which makes it possible to optimize the generally accepted resource allocation, significantly reducing the amount of the resources. Practical relevance: The presented models and the method of their analysis can be used for efficient planning and allocation of resources for data-intensive systems.

Fair and profitable: How pricing and lead-time quotation policies can help

In this paper, we propose four policies to serve price and lead-time sensitive customers with a single type of product produced in an M/GI/1 type make-to-stock queueing system. The policies are developed to observe certain principles of fairness: if a customer is quoted a longer lead-time, she/he must be charged a lower price and a significant proportion of the deliveries has to be made during the quoted lead times. Although handling non-exponential service times in this setting presents difficulties, our analysis of the proposed policies is exact. Two of the policies operate with static prices, while one of the two dynamic pricing policies also quotes dynamic lead times. By construction of the policies, we show that the dynamic pricing policies are more profitable. Numerical examples bring additional light by showing that in small markets with oversensitive customers, dynamic policies can be profitable while static pricing policies can fail. In a larger market, a simple dynamic policy charging two prices depending on the stock availability can be a reasonable compromise. Dynamic policies tend to suffer less against high production time variability as well.

Mathematical Model of Call Center in the Form of Multi-Server Queueing System

The paper considers the model of a call center in the form of a multi-server queueing system with Poisson arrivals and an unlimited waiting area. In the model under consideration, incoming calls do not differ in terms of service conditions, requested service, and interarrival periods. It is assumed that an incoming call can use any free server and they are all identical in terms of capabilities and quality. The goal problem is to find the stationary distribution of the number of calls in the system for an arbitrary recurrent service. This will allow us to evaluate the performance measures of such systems and solve various optimization problems for them. Considering models with non-exponential service times provides solutions for a wide class of mathematical models, making the results more adequate for real call centers. The solution is based on the approximation of the given distribution function of the service time by the hyperexponential distribution function. Therefore, first, the problem of studying a system with hyperexponential service is solved using the matrix-geometric method. Further, on the basis of this result, an approximation of the stationary distribution of the number of calls in a multi-server system with an arbitrary distribution function of the service time is constructed. Various issues in the application of this approximation are considered, and its accuracy is analyzed based on comparison with the known analytical result for a particular case, as well as with the results of the simulation.

Age of Information in G/G/1/1 Systems: Age Expressions, Bounds, Special Cases, and Optimization

We consider the average age of information in G/G/1/1 systems under two service discipline models. In the first model, if a new update arrives when the service is busy, it is blocked; in the second model, a new update preempts the current update in service. For the blocking model, we first derive an exact age expression for G/G/1/1 systems. Then, using the age expression for G/G/1/1 systems, we calculate average age expressions for special cases, i.e., M/G/1/1 and G/M/1/1 systems. We observe that deterministic interarrivals minimize the average age of G/M/1/1 systems for a given mean interarrival time. Next, for the preemption in service model, we first derive an exact average age expression for G/G/1/1 systems. Then, similar to blocking discipline, using the age expression for G/G/1/1 systems, we calculate average age expressions for special cases, i.e., M/G/1/1 and G/M/1/1 systems. Average age for G/M/1/1 can be written as a summation of two terms, the first of which depends only on the first and second moments of interarrival times and the second of which depends only on the service rate. In other words, interarrival and service times are decoupled. We prove that deterministic interarrivals are optimum for G/M/1/1 systems for a given mean interarrival time. On the other hand, we observe for non-exponential service times that the optimal distribution of interarrival times depends on the relative values of the mean interarrival time and the mean service time. Finally, we propose a simple to calculate upper bound to the average age for the preemption in service discipline.

On two product form modifications for finite overflow systems

Overflow mechanisms can be found in a variety of queueing models. This paper studies a simple and generic overflow system that allows the service times to be both job type and station dependent. This system does not exhibit a product form. To justify simple product form computations, two product form modifications are given, as by a so-called call packing principle and by a stop protocol. The provided proofs are self-contained and straightforward for the exponential case and of merit by itself. Next, it is numerically studied whether and when, or under which conditions, the modifications lead to a reasonable approximation of the blocking probability, if not an ordering. The numerical results indicate that call packing provides a rather accurate approximation when the overflow station is not heavily utilized. Moreover, when overflowed jobs have an equal or faster service rate, the approximation is consistently found to be pessimistic, which can be useful for practical purposes. The stop protocol, in contrast, appears to be less accurate for most natural situations. Nevertheless, for an extreme situation the order might change. In addition, for the stop protocol the product form is proven to be insensitive (i.e. to also apply for arbitrary non-exponential service times). For call packing, this numerically appears not to be the case, as of interest by itself. However, from a practical viewpoint the sensitivity seems light. The results are intriguing for both theoretical and practical further research.

Estimating the Response Time of a Cloud Computing System with the Help of Neural Networks

The article presents a new approach to assessing the average response time of a cloud computing system and its dispersion. A fork-join system or a system with request splitting was chosen as a queuing model, and artificial neural networks were used as a method for estimating a variable of interest. The analysis showed that the estimates obtained were more accurate than those previously known. Besides, the proposed approach allows expanding the analysis of the cloud system to the case of a model with a non-Poisson input stream and non-exponential service time, as well as obtaining estimates for a larger number of performance indicators of the cloud system, which was not previously possible.

Cooperative game in parallel service systems with nonexponential service times

Cooperative game in parallel service systems with nonexponential service times

Infinite-server queues with Hawkes input

AbstractIn this paper we study the number of customers in infinite-server queues with a self-exciting (Hawkes) arrival process. Initially we assume that service requirements are exponentially distributed and that the Hawkes arrival process is of a Markovian nature. We obtain a system of differential equations that characterizes the joint distribution of the arrival intensity and the number of customers. Moreover, we provide a recursive procedure that explicitly identifies (transient and stationary) moments. Subsequently, we allow for non-Markovian Hawkes arrival processes and nonexponential service times. By viewing the Hawkes process as a branching process, we find that the probability generating function of the number of customers in the system can be expressed in terms of the solution of a fixed-point equation. We also include various asymptotic results: we derive the tail of the distribution of the number of customers for the case that the intensity jumps of the Hawkes process are heavy tailed, and we consider a heavy-traffic regime. We conclude by discussing how our results can be used computationally and by verifying the numerical results via simulations.

Many-server Gaussian limits for overloaded non-Markovian queues with customer abandonment

Extending Ward Whitt’s pioneering work “Fluid Models for Multiserver Queues with Abandonments, Operations Research, 54(1) 37–54, 2006,” this paper establishes a many-server heavy-traffic functional central limit theorem for the overloaded $$G{/}GI{/}n+GI$$ queue with stationary arrivals, nonexponential service times, n identical servers, and nonexponential patience times. Process-level convergence to non-Markovian Gaussian limits is established as the number of servers goes to infinity for key performance processes such as the waiting times, queue length, abandonment and departure processes. Analytic formulas are developed to characterize the distributions of these Gaussian limits.

Staffing to Stabilize the Tail Probability of Delay in Service Systems with Time-Varying Demand

Analytic formulas are developed to set the time-dependent number of servers to stabilize the tail probability of customer waiting times for the Gt/GI/st + GI queueing model, which has a nonstationary non-Poisson arrival process (the Gt), nonexponential service times (the first GI), and allows customer abandonment according to a nonexponential patience distribution (the +GI). Specifically, for any delay target w > 0 and probability target α ∈ (0, 1), we determine appropriate staffing levels (the st) so that the time-varying probability that the waiting time exceeds a maximum acceptable value w is stabilized at α at all times. In addition, effective approximating formulas are provided for other important performance functions such as the probabilities of delay and abandonment, and the means of delay and queue length. Many-server heavy-traffic limit theorems in the efficiency-driven regime are developed to show that (i) the proposed staffing function achieves the goal asymptotically as the scale increases, and (ii) the proposed approximating formulas for other performance measures are asymptotically accurate as the scale increases. Extensive simulations show that both the staffing functions and the performance approximations are effective, even for smaller systems having an average of three servers. The e-companion is available at https://doi.org/10.1287/opre.2017.1678 .

Stationary queue length of a single-server queue with negative arrivals and nonexponential service time distributions

In this paper, a single-server queue with negative customers is considered. The arrival of a negative customer will remove one positive customer that is being served, if any is present. An alternative approach will be introduced to derive a set of equations which will be solved to obtain the stationary queue length distribution. We assume that the service time distribution tends to a constant asymptotic rate when time t goes to infinity. This assumption will allow for finding the stationary queue length of queueing systems with non-exponential service time distributions. Numerical examples for gamma distributed service time with fractional value of shape parameter will be presented in which the steady-state distribution of queue length with such service time distributions may not be easily computed by most of the existing analytical methods.

Coping with production time variability via dynamic lead-time quotation

In this paper, we propose two dynamic lead-time quotation policies in an M/GI/1 type make-to-stock queueing system serving lead-time sensitive customers with a single type of product. Incorporating non-exponential service times in an exact method for make-to-stock queues is usually deemed difficult. Our analysis of the proposed policies is exact and requires the numerical inversion of the Laplace transform of the sojourn time of an order to be placed. The first policy assures that the long-run probability of delivering the product within the quoted lead-time is the same for all backlogged customers. The second policy is a refinement of the first which improves the profitability if customers are oversensitive to even short delays in delivery. Numerical results show that both policies perform close to the optimal policy that was characterized only for exponential service times. The new insight gained is that the worsening impact of the production time variability, which is felt significantly in systems accepting all customers by quoting zero lead times, decreases when dynamic lead-time quotation policies are employed.

Block-structured supermarket models

Supermarket models are a class of parallel queueing networks with an adaptive control scheme that play a key role in the study of resource management of, such as, computer networks, manufacturing systems and transportation networks. When the arrival processes are non-Poisson and the service times are non-exponential, analysis of such a supermarket model is always limited, interesting, and challenging. This paper describes a supermarket model with non-Poisson inputs: Markovian Arrival Processes (MAPs) and with non-exponential service times: Phase-type (PH) distributions, and provides a generalized matrix-analytic method which is first combined with the operator semigroup and the mean-field limit. When discussing such a more general supermarket model, this paper makes some new results and advances as follows: (1) Providing a detailed probability analysis for setting up an infinite-dimensional system of differential vector equations satisfied by the expected fraction vector, where the invariance of environment factors is given as an important result. (2) Introducing the phase-type structure to the operator semigroup and to the mean-field limit, and a Lipschitz condition can be established by means of a unified matrix-differential algorithm. (3) The matrix-analytic method is used to compute the fixed point which leads to performance computation of this system. Finally, we use some numerical examples to illustrate how the performance measures of this supermarket model depend on the non-Poisson inputs and on the non-exponential service times. Thus the results of this paper give new highlight on understanding influence of non-Poisson inputs and of non-exponential service times on performance measures of more general supermarket models.

Algorithms for Time-Varying Networks of Many-Server Fluid Queues

Motivated by large-scale service systems with network structure, we introduced in a previous paper a time-varying open network of many-server fluid queues with customer abandonment from each queue and time-varying proportional routing among the queues, and showed how performance functions can be determined. The deterministic fluid model serves as an approximation for the corresponding non-Markovian stochastic network of many-server queues with Markovian routing, experiencing periods of overloading at the queues. In this paper we develop a new algorithm for the previous model and generalize the model to include non-exponential service-time distributions. In this paper we report results of implementing the algorithms and studying their computational complexity. We also conduct simulation experiments to confirm that the algorithms are effective in computing the performance functions and that these performance functions provide useful approximations for the corresponding stochastic models.

Decay of tails at equilibrium for FIFO join the shortest queue networks

In join the shortest queue networks, incoming jobs are assigned to the shortest queue from among a randomly chosen subset of $D$ queues, in a system of $N$ queues; after completion of service at its queue, a job leaves the network. We also assume that jobs arrive into the system according to a rate-$\alpha N$ Poisson process, $\alpha<1$, with rate-1 service at each queue. When the service at queues is exponentially distributed, it was shown in Vvedenskaya et al. [Probl. Inf. Transm. 32 (1996) 15-29] that the tail of the equilibrium queue size decays doubly exponentially in the limit as $N\rightarrow\infty$. This is a substantial improvement over the case D=1, where the queue size decays exponentially. The reasoning in [Probl. Inf. Transm. 32 (1996) 15-29] does not easily generalize to jobs with nonexponential service time distributions. A modularized program for treating general service time distributions was introduced in Bramson et al. [In Proc. ACM SIGMETRICS (2010) 275-286]. The program relies on an ansatz that asserts, in equilibrium, any fixed number of queues become independent of one another as $N\rightarrow\infty$. This ansatz was demonstrated in several settings in Bramson et al. [Queueing Syst. 71 (2012) 247-292], including for networks where the service discipline is FIFO and the service time distribution has a decreasing hazard rate. In this article, we investigate the limiting behavior, as $N\rightarrow \infty$, of the equilibrium at a queue when the service discipline is FIFO and the service time distribution has a power law with a given exponent $-\beta$, for $\beta>1$. We show under the above ansatz that, as $N\rightarrow\infty$, the tail of the equilibrium queue size exhibits a wide range of behavior depending on the relationship between $\beta$ and $D$. In particular, if $\beta>D/(D-1)$, the tail is doubly exponential and, if $\beta<D/(D-1)$, the tail has a power law. When $\beta=D/(D-1)$, the tail is exponentially distributed.

Asymptotic independence of queues under randomized load balancing

Randomized load balancing greatly improves the sharing of resources while being simple to implement. In one such model, jobs arrive according to a rate-?N Poisson process, with ?<1, in a system of N rate-1 exponential server queues. In Vvedenskaya et al. (Probl. Inf. Transm. 32:15---29, 1996), it was shown that when each arriving job is assigned to the shortest of D, D?2, randomly chosen queues, the equilibrium queue sizes decay doubly exponentially in the limit as N??. This is a substantial improvement over the case D=1, where queue sizes decay exponentially. The reasoning in Vvedenskaya et al. (Probl. Inf. Transm. 32:15---29, 1996) does not easily generalize to jobs with nonexponential service time distributions. A modularized program for treating randomized load balancing problems with general service time distributions was introduced in Bramson et al. (Proc. ACM SIGMETRICS, pp. 275---286, 2010). The program relies on an ansatz that asserts that, for a randomized load balancing scheme in equilibrium, any fixed number of queues become independent of one another as N??. This allows computation of queue size distributions and other performance measures of interest. In this article, we demonstrate the ansatz in several settings. We consider the least loaded balancing problem, where an arriving job is assigned to the queue with the smallest workload. We also consider the more difficult problem, where an arriving job is assigned to the queue with the fewest jobs, and demonstrate the ansatz when the service discipline is FIFO and the service time distribution has a decreasing hazard rate. Last, we show the ansatz always holds for a sufficiently small arrival rate, as long as the service distribution has 2 moments.

Analysis of Cellular Wireless Networks with Non-Classical Traffic and Channel Reservation

This paper describes a traffic model for cellular wireless network with non-classical traffic (i.e., gamma inter-arrivals and general service times). Empirical studies have shown that electromagnetic propagation condition, irregular users’ behaviours, management policies, equipment failures, charging capacity of battery and so on have significant influence in call drops. Hence, the exponential assumption for both the inter-arrival and service times does not always hold true, as it does not represent the real dynamics involved in an operational network. The proposed Ek/G/C model takes into account the non-exponential inter-arrival time and service time distributions and gives a more pragmatic model for future cellular wireless networks. The performance of the proposed model of the cellular wireless network has been re-evaluated with channel reservation using both classical and non-classical traffic. The results show that non-classical model gives better approximation of the system performance in terms of the blocking probabilities.

Performance modelling of autonomous vehicle storage and retrieval systems with generally distributed service times

This paper proposes extensions of system conceptualisation models for autonomous vehicle storage and retrieval systems (AVS/RSs) to deal with non-exponential lift service times when the number of lifts and vehicles are both design variables. These extensions are based on an adaptation of a computational algorithm for estimating the probability distribution of queue lengths in capacitated, non-Poisson queuing systems. Simulation-based validation studies suggest that the proposed model extensions yield estimation errors comparable to those observed in earlier studies restricted to cases where the number of vehicle and lifts is treated as a single design variable.

Two-parameter heavy-traffic limits for infinite-server queues

In order to obtain Markov heavy-traffic approximations for infinite-server queues with general non-exponential service-time distributions and general arrival processes, possibly with time-varying arrival rates, we establish heavy-traffic limits for two-parameter stochastic processes. We consider the random variables Q e (t,y) and Q r (t,y) representing the number of customers in the system at time t that have elapsed service times less than or equal to time y, or residual service times strictly greater than y. We also consider W r (t,y) representing the total amount of work in service time remaining to be done at time t+y for customers in the system at time t. The two-parameter stochastic-process limits in the space D([0,∞),D) of D-valued functions in D draw on, and extend, previous heavy-traffic limits by Glynn and Whitt (Adv. Appl. Probab. 23, 188–209, 1991), where the case of discrete service-time distributions was treated, and Krichagina and Puhalskii (Queueing Syst. 25, 235–280, 1997), where it was shown that the variability of service times is captured by the Kiefer process with second argument set equal to the service-time c.d.f.