A multivariate Markov chain model for categorical data sequences and its applications in demand predictions

Markov models are commonly used in modelling many practicalsystems such as telecommunication systems, manufacturing systemsand inventory systems. In this paper we propose a multivariateMarkov chain model for modelling multiple categorical data sequences.We develop efficient estimation methods for the model parameters.We then apply the model and method to demand predictions fora soft-drink company in Hong Kong.     

Inventory control for point-of-use locations in hospitals

Most inventory management systems at hospital departments are characterised by lost sales, periodic reviews with short lead times, and limited storage capacity. We develop two types of exact models that deal with all these characteristics. In a capacity model, the service level is maximised subject to a capacity restriction, and in a service model the required capacity is minimised subject to a service level restriction. We also formulate approximation models applicable for any lost-sales inventory system (cost objective, no lead time restrictions etc). For the capacity model, we develop a simple inventory rule to set the reorder levels and order quantities. Numerical results for this inventory rule show an average deviation of 1% from the optimal service levels. We also embed the single-item models in a multi-item system. Furthermore, we compare the performance of fixed order size replenishment policies and (R,?s,?S) policies.     

The finite capacity multi-server queue with inhomogeneous arrival rate and discrete service time distribution — and its application to continuous service time problems

The finite capacity multi-server queueing model with inhomogeneous arrival rate and discrete service time distribution is developed. The system is formulated as an inhomogeneous Markov chain in discrete time. An algorithm is described to solve the model numerically. A method is then proposed for using this model to approximate the time dependent behaviour of multi-server queues with inhomogeneous arrival rate and continuous service time distribution. Empirical results are presented to show that this approximation will produce results that are accurate enough for most practical purposes.     

Analysis of the BMAP/PH/N queueing system with backup servers

We consider a novel multi-server queueing system that is potentially useful for optimizing real-world systems, in which the objectives of high performance and low power consumption are conflicting. The queueing model is formulated and investigated under the assumption that an arrival flow is defined by a batch Markovian arrival process and random values characterizing customer processing have the phase-type distribution. If the service time of some customer by a server exceeds a certain random bound, this server receives help from a so-called backup server from a finite pool of backup servers. The behavior of the system is described by a quite complicated multi-dimensional continuous-time Markov chain that is successfully analyzed in this paper. Examples of the potential use of the obtained results in managerial decisions are presented.     

A possibilistic decision model for new product supply chain design

This paper models supply chain (SC) uncertainties by fuzzy sets and develops a possibilistic SC configuration model for new products with unreliable or unavailable SC statistical data. The supply chain is modeled as a network of stages. Each stage may have one or more options characterized by the cost and lead-time required to fulfill required functions and may hold safety stock to prevent an inventory shortage. The objective is to determine the option and inventory policy for each stage to minimize the total SC cost and maximize the possibility of fulfilling the target service level. A fuzzy SC model is developed to evaluate the performance of the entire SC and a genetic algorithm approach is applied to determine near-optimal solutions. The results obtained show that the proposed approach allows decision makers to perform trade-off analysis among customer service levels, product cost, and inventory investment depending on their risk attitude. It also provides an alternative tool to evaluate and improve SC configuration decisions in an uncertain SC environment.     

Polling systems in the critical regime

We study polling systems with Poisson arrival streams and exhaustive service discipline in the critical load case ρ=1. We show that the system may exhibit null-recurrence or transient behavior and give the corresponding conditions in terms of the first two moments of the distribution of service and switching times. For models with two nodes we establish existence of moments of small orders for the return time to the origin of the corresponding embedded Markov chain in the case of null-recurrence.     

A Markov-modulated M/G/1 queue II: Busy period and time for buffer overflow

We consider an M/G/1 queue where the arrival and service processes are modulated by a two state Markov chain. We assume that the arrival rate, service time density and the rates at which the Markov chain switches its state, are functions of the total unfinished work (buffer content) in the queue. We compute asymptotic approximations to performance measures such as the mean residual busy period, mean length of a busy period, and the mean time to reach capacity.This research was supported in part by NSF Grants DMS-84-06110, DMS-85-01535 and DMS-86-20267, and grants from the U.S. Israel Binational Science Foundation and the Israel Academy of Sciences.     

Integrated capacity and inventory management with capacity acquisition lead times

We model a make-to-stock production system that utilizes permanent and contingent capacity to meet non-stationary stochastic demand, where a constant lead time is associated with the acquisition of contingent capacity. We determine the structure of the optimal solution concerning both the operational decisions of integrated inventory and flexible capacity management, and the tactical decision of determining the optimal permanent capacity level. Furthermore, we show that the inventory (either before or after production), the pipeline contingent capacity, the contingent capacity to be ordered, and the permanent capacity are economic substitutes. We also show that the stochastic demand variable and the optimal contingent capacity acquisition decisions are economic complements. Finally, we perform numerical experiments to evaluate the value of utilizing contingent capacity and to study the effects of capacity acquisition lead time, providing useful managerial insights.     

Coordination mechanism for capacity reservation by considering production time,production rate and order quantity

In this paper we study the coordination of a dyadic supply chain producing a high-tech product by contracts. The product has a short life cycle and the buyer faces stochastic demands during the selling period. We consider the production time, which causes the inventory costs on supplier’s side. As the supplier builds production capacity in advance, the production rate is limited to the capacity created during the production time. In addition, we take into account the inventory cost and operational cost for the buyer. We examine the model under both full information and partial information updating situations, and propose a coordinating contract for each case. Our analysis includes the study of members’ decisions under both forced and voluntary compliance regimes. Numerical results are presented to provide more insights into the models developed and the mechanisms proposed.     

Flow control in capacity-constrained queueing systems with non-stationary arrivals

In this paper, we present a flow control policy to address the problem of service capacity constraints in queueing systems that face non-stationary customer arrival rates. The studied system concerns commercial trucks arriving at international border checkpoints for processing. Based on an extensive range of numerical examples, we quantify how the flow control policy of bringing arrival rates more in line with available capacity affects customer-oriented metrics and measures of efficiency in service capacity resources. We also quantify the expectations that the policy would place on arriving customers. How these quantitative outcomes are influenced by contextual parameters such as the pattern of customer arrivals and the tightness of capacity constraints is also clarified.     

Transient and Stationary Distributions for the GI/G/k Queue with Lebesgue-Dominated Inter-Arrival Time Distribution

In this paper, the multi-server queue with general service time distribution and Lebesgue-dominated iid inter-arival times is analyzed. This is done by introducing auxiliary variables for the remaining service times and then examining the embedded Markov chain at arrival instants. The concept of piecewise-deterministic Markov processes is applied to model the inter-arrival behaviour. It turns out that the transition probability kernel of the embedded Markov chain at arrival instants has the form of a lower Hessenberg matrix and hence admits an operator–geometric stationary distribution. Thus it is shown that matrix–analytical methods can be extended to provide a modeling tool even for the general multi-server queue.     

A two-stage queueing network on form postponement supply chain with correlated demands

To ease the conflict between quick response and product variety, more and more business models are developed in supply chains. Among these, the form postponement (FP) strategy is an efficient tool and has been widely adopted. To the supply chain with FP strategy, the design mostly involves two problems: determination of customer order decoupling point (CODP) position and semi-finished product inventory control. In this paper, we develop a two-stage tandem queueing network with MAP arrival to address this issue. Particularly, we introduce a Markov arrival process (MAP) to characterize the correlation of the demand. By using of matrix geometric method, we derive several performance measure of the supply chain, such as inventory level and unfill rate. Our numerical examples show that both the variance and the correlation coefficient of the demand lead to more delayed CODP position and more total cost.     

Counterexamples for comparisons of queues with finite waiting rooms

We constructG/G/1/k queueing models that fail to have anticipated monotonicity properties with respect to the capacityk. In one model the long-run average number of customers in the system is arbitrarily close to the capacityk, but it decreases to an arbitrarily small value when the capacity is increased. In another model the throughput is arbitrarily close to the arrival rate when the capacity isk, but the throughput decreases to an arbitrarily small value when the capacity is increased. These examples involving non i.i.d. service times, which are associated with external arrivals instead of being assigned when service begins, show that stochastic assumptions and arguments involving more than direct sample-path comparisons are essential for obtaining useful bounds and positive comparison results.     

Analysis of a Discrete-Time Finite-Capacity Single-Server Queue with Markovian-Arrival Process and Random-Bulk-Service: D-MAP/G Y /1/M

In this article, we consider a single-server, finite-capacity queue with random bulk service rule where customers arrive according to a discrete-time Markovian arrival process (D-MAP). The model is denoted by D-MAP/G ^Y /1/M where server capacity (bulk size for service) is determined by a random variable Y at the starting point of services. A simple analysis of this model is given using the embedded Markov chain technique and the concept of the mean sojourn time of the phase of underlying Markov chain of D-MAP. A complete solution to the distribution of the number of customers in the D-MAP/G ^Y /1/M queue, some computational results, and performance measures such as the average number of customers in the queue and the loss probability are presented.     

On distributed scheduling with heterogeneously delayed network-state information

We study the problem of distributed scheduling in wireless networks, where each node makes individual scheduling decisions based on heterogeneously delayed network state information (NSI). This leads to inconsistency in the views of the network across nodes, which, coupled with interference, makes it challenging to schedule for high throughputs. We characterize the network throughput region for this setup, and develop optimal scheduling policies to achieve the same. Our scheduling policies have a threshold-based structure and, moreover, require the nodes to use only the “smallest critical subset” of the available delayed NSI to make decisions. In addition, using Markov chain mixing techniques, we quantify the impact of delayed NSI on the throughput region. This not only highlights the value of extra NSI for scheduling, but also characterizes the loss in throughput incurred by lower complexity scheduling policies which use homogeneously delayed NSI.     

Performance Evaluation Models for Single-item Periodic Pull Production Systems

A number of pull production systems reported in the literature are found to be equivalent to a tandemqueue so that existing accurate tandem-queue approximation methods can be used to evaluate such systems. In this study, we consider developing an exact performance evaluation model for a non-tandemqueue equivalent pull production system using discrete-time Markov processes. It is a periodically controlled serial production system in which a single-item is processed at each stage with an exponential processing time in order to satisfy the Poisson finished product demand. The selected performance measures are throughput, inventory levels, machine utilizations and service level of the system. For large systems, which are difficult to evaluate exactly because of large state-spaces involved, we also propose a computationally feasible approximate decomposition technique together with some numerical experimentations.     

An integrated model for warehouse and inventory planning

The purpose of this article is to evaluate the value of integrating tactical warehouse and inventory decisions. Therefore, a global warehouse and inventory model is presented and solved. In order to solve this mathematical model, two solution methodologies are developed which offer different level of integration of warehouse and inventory decisions. Computational tests are performed on a real world database using multiple scenarios differing by the warehouse capacity limits and the warehouse and inventory costs. Our observation is that the total cost of the inventory and warehouse systems can be reduced drastically by taking into account the warehouse capacity restrictions in the inventory planning decisions, in an aggregate way. Moreover additional inventory and warehouse savings can be achieved by using more sophisticated integration methods for inventory and warehouse decisions.     

The impact of forecasting model selection on the value of information sharing in a supply chain

This paper presents a study on the impact of forecasting model selection on the value of information sharing in a supply chain with one capacitated supplier and multiple retailers. Using a computer simulation model, this study examines demand forecasting and inventory replenishment decisions by the retailers, and production decisions by the supplier under different demand patterns and capacity tightness. Analyses of the simulation output indicate that the selection of the forecasting model significantly influences the performance of the supply chain and the value of information sharing. Furthermore, demand patterns faced by retailers and capacity tightness faced by the supplier also significantly influence the value of information sharing. The result also shows that substantial cost savings can be realized through information sharing and thus help to motivate trading partners to share information in the supply chain. The findings can also help supply chain managers select suitable forecasting models to improve supply chain performance.     

Monte Carlo Maximum Likelihood in Model-Based Geostatistics

When using a model-based approach to geostatistical problems, often, due to the complexity of the models, inference relies on Markov chain Monte Carlo methods. This article focuses on the generalized linear spatial models, and demonstrates that parameter estimation and model selection using Markov chain Monte Carlo maximum likelihood is a feasible and very useful technique. A dataset of radionuclide concentrations on Rongelap Island is used to illustrate the techniques. For this dataset we demonstrate that the log-link function is not a good choice, and that there exists additional nonspatial variation which cannot be attributed to the Poisson error distribution. We also show that the interpretation of this additional variation as either micro-scale variation or measurement error has a significant impact on predictions. The techniques presented in this article would also be useful for other types of geostatistical models.     

Waiting times of a finite-capacity multi-server model with non-preemptive priorities

We consider a non-preemptive head of the line multi-server priority model with finite capacity. The arrival processes of the different priority classes are independent Poisson processes. The service times are exponentially distributed and identical for the different priority classes. The model is described by a homogeneous continuous-time Markov chain. For the two-class model we derive an explicit representation of its steady-state distribution. Applying matrix-analytic methods we calculate the Laplace-Stieltjes Transform of the actual waiting time of each priority class of a p-class system.