Jeffrey P. Kharoufeh

• Ph.D.
• Professor at Clemson University

We propose a new nonconvex regularization scheme to improve the performance of the stochastic dual dynamic programming (SDDP) algorithm for solving large-scale multistage stochastic programs. Specifically, we use a class of nonconvex regularization functions, namely folded concave penalty functions, to improve solution quality and the convergence r...

We consider the problem of optimally maintaining an offshore wind farm in which major components progressively degrade over time due to normal usage and exposure to a randomly varying environment. The turbines exhibit both economic and stochastic dependence due to shared maintenance setup costs and their common environment. Our aim is to identify o...

The primary aim of this chapter is to provide an overview of some of the most important stochastic models that are widely employed by the industrial and systems engineer. This survey is by no means exhaustive but is intended to highlight the most prevalent models and their use in important application areas. While stochastic computer simulation mod...

We examine necessary and sufficient conditions for recurrence and positive recurrence of a class of irreducible, level-dependent quasi-birth-and-death (LDQBD) processes with a block tridiagonal structure that exhibits asymptotic convergence in the rows as the level tends to infinity. These conditions are obtained by exploiting a multi-dimensional L...

We consider the problem of optimally maintaining a stochastically degrading, single‐unit system using heterogeneous spares of varying quality. The system's failures are unannounced; therefore, it is inspected periodically to determine its status (functioning or failed). The system continues in operation until it is either preventively or corrective...

We formulate an optimal stopping problem for a variant of Shepp’s urn model in which it is possible to sample more than one item at each stage. Using a Markov decision process model, we establish monotonicity of the optimal value function and show that the optimal policy is a monotone threshold policy that prescribes either not sampling, or samplin...

We consider the problem of optimally replacing multiple stochastically degrading systems using condition-based maintenance. Each system degrades continuously at a rate that is governed by the current state of the environment, and each fails once its own cumulative degradation threshold is reached. The objective is to minimize the sum of the expecte...

We consider the problem of dynamically controlling a 2-bus energy distribution network with energy storage capabilities. An operator seeks to dynamically adjust the amount of energy to charge to, or discharge from, energy storage devices in response to randomly-evolving demand, renewable supply and prices. The objective is to minimize the expected...

A microgrid is a small-scale version of a centralized power grid that generates, distributes and regulates electricity flow to local entities using distributed generation and the main grid. Distributed energy storage systems can be used to mitigate adverse effects of intermittent renewable sources in a microgrid in which operators dynamically adjus...

Linear programming (LP) formulations are often employed to solve stationary, infinite-horizon Markov decision process (MDP) models. We present an LP approach to solving non-stationary, finite-horizon MDP models that can potentially overcome the computational challenges of standard MDP solution procedures. Specifically, we establish the existence of...

We consider the integrated problem of optimally maintaining an imperfect, deteriorating sensor and the safety-critical system it monitors. The sensor's costless observations of the binary state of the system become less informative over time. A costly full inspection may be conducted to perfectly discern the state of the system, after which the sys...

Consider a two-station, heterogeneous parallel queueing system in which each station operates as an independent M / M /1 queue with its own infinite-capacity buffer. The input to the system is a Poisson process that splits among the two stations according to a Bernoulli splitting mechanism. However, upon arrival, a strategic customer initially join...

Motivated by wind energy applications, we consider the problem of optimally replacing a stochastically degrading component that resides and operates in a partially observable environment. The component's rate of degradation is modulated by the stochastic environment process, and the component fails when it is accumulated degradation first reaches a...

This article presents a methodology for modeling degradation signals from components functioning under dynamically evolving environment conditions. In situ sensor signals related to the degradation process are utilized as well as the environment conditions, to predict and update, in real-time, the distribution of a component’s residual lifetime. Th...

We consider the problem of optimally locating application virtualization resources among a network of military treatment facilities within the Military Health System (MHS). Application virtualization consolidates computer application processing and administration in a centralized data center and delivers a virtual image of the application over the...

We consider the problem of controlling a two-server Markovian queueing system with heterogeneous servers. The servers are differentiated by their service rates and reliability attributes (i.e., the slower server is perfectly reliable, whereas the faster server is subject to random failures). The aim is to dynamically route customers at arrival, ser...

This article presents a switching diffusion model for estimating the useful lifetime of a component that operates in a randomly varying environment. The component's degradation process is unobservable; therefore, a signal of degradation is observed to estimate the environment parameters using a Markov chain Monte Carlo statistical procedure. These...

We consider the problem of maximizing the lifetime of a query-based wireless sensor network in which all of the sensor nodes are both producers and consumers of network resources. Of particular concern is the problem of selecting a common transmission range for all of the sensor nodes, the resource replication level (or time-to-live counter), and t...

This paper presents a preliminary stochastic optimization model to prescribe energy storage strategies within a smart electric power microgrid. A microgrid is a small-scale version of a centralized electric power grid that generates, distributes and regulates electricity flow to local residential consumers using distributed renewable sources, as we...

Efrosinin (Ann Oper Res 202:75–102, 2013) examined the optimal allocation of customers in an \(M/M/2\) queueing system with heterogeneous servers differentiated by their service rates and reliability attributes. Specifically, the faster server is subject to partial or complete failures, and the slower server is perfectly reliable. The objective is...

We present two stochastic failure models for the reliability evaluation of manufacturing equipment that degrades due to its complex operating environment. The first model examines the case when the environment is a temporally nonhomogeneous continuous-time Markov chain, and the second assumes the environment is a temporally homogeneous semi-Markov...

We present a queueing-theoretic framework to evaluate the performance of large-scale, query-based wireless sensor networks whose nodes detect and advertise significant events that are useful for only a limited time; queries generated by sensor nodes are also time-limited. The main performance parameter is the steady state proportion of generated qu...

Intensive care units (ICUs) are limited-capacity, resource-intensive wards in a hospital designed to provide continuously monitored, intensive care and temporary support to critically-ill patients with a broad range of health conditions. Therefore, their efficient operation and management are critical to providing quality care to the most severely...

Endowing conventional microcrystalline materials with nanometer-scale grains at the surfaces can offer enhanced mechanical properties, including improved wear, fatigue, and friction properties, while simultaneously enabling useful functionalizations with regard to biocompatibility, osseointegration, electrochemical performance, etc. To inherit such...

We examine the problem of optimally maintaining a stochastically degrading system using preventive and reactive replacements. The system's rate of degradation is modulated by an exogenous stochastic environment process, and the system fails when its cumulative degradation level first reaches a fixed deterministic threshold. The objective is to mini...

We examine an M/M/1 retrial queue with an unreliable server whose arrival, service, failure, repair, and retrial rates are all modulated by an exogenous random environment. Provided are conditions for stability, the (approximate) orbit size distribution, and mean queueing performance measures which are obtained via matrix-analytic methods. Addition...

We consider the problem of estimating the lifetime distribution of a component whose degradation is driven by an exogenous environment process. The environment is assumed to be unobservable; therefore, its order and transition rates must be inferred from a sequence of degradation observations. We present and assess the effectiveness of a Hidden Mar...

This article defines and describes the level-dependent quasi-birth-and-death (LDQBD) process, a generalization of the homogeneous (or level-independent) quasi-birth-and-death (QBD) process. Like its level-independent counterpart, the LDQBD process is a bivariate Markov process whose transition probability matrix (or infinitesimal generator matrix)...

This article describes the M/G/s/s queueing system, also known as the Erlang loss model. Customers arrive to this stochastic service system according to a homogeneous Poisson process and request service from one of s parallel servers (s ≥ 1). The service times are assumed to be independent and identically distributed with a finite mean. Because the...

This article describes the batch Markovian arrival process (BMAP), a point process that is characterized by Markov-modulated batch arrivals of random size. The BMAP is a gener-alization of many well-known processes including the Markovian arrival process (MAP), the Poisson process, and the Markov-modulated Poisson process. It provides a common fram...

Recently, Sherman et al. [14] analyzed an M/G/1 retrial queueing model in which customers are forced to retry their service if interrupted by a server failure. Using classical techniques, they provided a stability analysis, queue length distributions, key performance parameters, and stochastic decomposition results. We analyze the system under a st...

This article describes the batch Markovian arrival process (BMAP), a point process that is characterized by Markov-modulated batch arrivals of random size. The BMAP is a generalization of many well-known processes including the Markovian arrival process (MAP), the Poisson process, and the Markov-modulated Poisson process. It provides a common frame...

We consider the problem of optimally maintaining a periodically inspected system that deteriorates according to a discrete-time Markov process and has a limit on the number of repairs that can be performed before it must be replaced. After each inspection, a decision maker must decide whether to repair the system, replace it with a new one, or leav...

We consider the optimal replacement of a periodically inspected system under Markov deterioration that operates in a controlled environment. Provided are sufficient conditions that characterize an optimal control-limit replacement policy with respect to the system’s condition and its environment. The structure of the optimal policy is illustrated b...

This article presents hybrid, degradation-based reliability models for a single-unit system whose degradation is driven by a semi-Markov environment. The primary objective is to develop a mathematical framework and associated computational techniques that unite environmental data and stochastic failure models to assess the current or future health...

We consider the problem of optimally selecting and specifying satellite payloads for inclusion on a sequence of satellite launches to maximize the total expected utility of a constellation. The satellite bus is constrained by energy, weight, volume, and cost limitations. We formulate the problem as a variant of the multidimensional knapsack problem...

We consider the problem of optimally timing the deployment of individual components in a multi-component system with the objective of maximizing the overall system utility over a finite planning horizon. The condition of each component, and its priority level, evolve stochastically over time while deployment decisions are subject to availability co...

We examine the problem of optimally maintaining a stochastically degrading system using preventive replacements. The system's rate of degradation is governed by an external environment, and the system fails once its cumulative degradation level reaches a fixed threshold. The objective is to minimize the total expected discounted cost of preventive...

We present transient and asymptotic reliability indices for a single-unit system that is subject to Markov-modulated shocks and wear. The transient results are derived from the (transform) solution of an integro-differential equation describing the joint distribution of the cumulative degradation process and the state of the modulating process. Add...

As a model for streaming multimedia applications, we study an unreliable retrial queue with inflnite-capacity orbit and normal queue for which the retrial rate and the server repair rate are controllable. Customers join the retrial orbit if and only if their service is interrupted by a server failure. Interrupted customers do not rejoin the normal...

We develop a queueing model for analyzing resource replication strategies in wireless sensor networks. The model can be used to minimize either the total transmission rate of the network (an energy-centric approach) or to ensure the proportion of query failures does not exceed a predetermined threshold (a failure-centric approach). The model explic...

We examine the performance of a random-walk search algorithm for wireless sensor networks when resources are subject to limited lifetimes and queries are constrained by application-specific deadlines. Specifically, via the time-to-live and transmission range parameters, we estimate the appropriate number of resource copies that must be created with...

The routine inspection of systems that do not exhibit obvious failures is essential to maintaining a high level of system availability in a variety of contexts including military equipment, manufacturing equipment and electrical systems. We formulate and solve an optimization model for selecting the inspection interval that maximizes the limiting a...

We analyze an M/M/1 retrial queue subject to breakdowns whose parameters are modulated by an external environment. We discuss stability conditions for this system and show that the joint process of the orbit size, environment, and server status is a tri-layered quasi-birth-and-death (QBD) process. Subsequently, we use matrix-geometric methods to ap...

We present a small-footprint search protocol designed to facilitate any-type queries for data content and services in large population, high-density wireless sensor networks. Our protocol, termed Trajectory-based Selective Broadcast Query (TSBQ), works in conjunction with time division multiple access- or schedule-based medium access control protoc...

We consider a two-server retrial queueing system in which both servers are subject to random failures that force interrupted customers to join the retrial orbit from which they persistently attempt to regain access to a server. The failure processes are mutually independent as are customer retrials. Using a phase-merging algorithm, we compute the a...

We analyze an unreliable M/M/1 retrial queue with infinite-capacity orbit and normal queue. Retrial customers do not rejoin the normal queue but repeatedly attempt to access the server at i.i.d. intervals until it is found functioning and idle. We provide stability conditions as well as several stochastic decomposability results.

The Bluetooth discovery process requires use of the inquiry substate which not only consumes significant power, but prevents normal data traffic flow and simultaneously acts as a potential noise source for neighboring networks. Therefore, the inquiry substate dwell time should be limited to discovering an acceptable number of neighboring devices. A...

We consider the problem of selecting an appropriate distortion function and associated parameters to account for rare but catastrophic events that may result from a shortfall of military or security capabilities. Additionally, we describe the means by which a decision maker may allocate resources among various risk-mitigating systems, subject to a...

We analyze a periodically inspected system with hidden failures in which the rate of wear is modulated by a continuous-time Markov chain and additional damage is induced by a Poisson shock process. We explicitly derive the system's lifetime distribution and mean time to failure, as well as the limiting average availability. The main results are ill...

We analyze a periodically inspected system with hidden failures in which the rate of wear is modulated by a continuous-time Markov chain and additional damage is induced by a Poisson shock process. We explicitly derive the system's lifetime distribution and mean time to failure, as well as the limiting average availability. The main results are ill...

We analyze an unreliable M/G/1 retrial queue with controllable repair and retrial rates whose orbit and normal queue both have an infinite waiting room. Presented are standard performance measures and an optimization problem to select the repair and retrial rates that minimize the long-run holding and service costs.

This article presents an analytical model for a few important operational characteristics of single-lane roundabouts in the United States. In particular, we obtain renewal-based analytical expressions for the mean and variance of the time required for entry into the circulating stream for an arbitrary vehicle occupying the first position of the app...

Abstract This paper presents a degradation-based procedure for estimating the full and residual lifetime distribution of a single-unit system subject to Markovian deterioration. The hybrid approach unites real degradation measures with analytical, stochastic failure models to numerically compute the distributions and their moments. Two distinct mod...

We present simplified analytical results for the numerical evaluation of failure time probabilities for a single-unit system whose cumulative wear over time depends on its external environment. The failure time distribution is derived as a one-dimensional Laplace–Stieltjes transform with respect to the temporal variable using a direct solution appr...

Surface-micromachined microelectromechanical systems (MEMS) are two dimensional in their “as fabricated” form. Surface-micromachined MEMS can be assembled after fabrication to realize systems with a more three-dimensional form and function. One reliable method of assembly, suited to commercial mass production, is to use the surface tension of micro...

We study the sensitivity of solutions of linear semidefinite programs under small arbitrary perturbations of the data. We present an elementary and self-contained proof of the differentiability of the solutions as functions of the perturbations, and ...

We analyze the moments of the random time required for a vehicle to traverse a transportation network link of arbitrary length when its speed is governed by a random environment. The problem is motivated by stochastic transportation network applications in which the estimation of travel time moments is of great importance. We analyze this random ti...

We present a stochastic model for the performance evaluation of a key phase in the military deployment process, joint reception, staging, onward movement, and integration (JRSOI). The process is modeled as an open, multiclass queueing network, wherein personnel and various classes of cargo are modeled as the flow entities and the phases of the proc...

We identify and analyze a specification-compatible method for implementing the Bluetooth inquiry function that simplifies the design while reducing the mean inquiry time. Using simulation data and independent empirical measurements, we verify a previously derived probability density function for the inquiry time of a Bluetooth device. Although the...

We derive an analytical expression for the cumulative distribution function of travel time for a vehicle traversing a freeway link of arbitrary length. The vehicle's speed is assumed to be modulated by a random environment that can be modeled as a stochastic process. We first present a partial differential equation (PDE) describing the travel time...

In this paper, we present a new integrated architecture and feature selection algorithm for radial basis neural networks (RBNNs). The objective is to apply the network in an iterative fashion to determine the final architecture and feature set used for the purpose of classification. Numerical experiments were conducted to compare the performance of...

The Bluetooth frequency hop sequence is uniformly distributed asymptotically but generally distributed otherwise. We present an upper bound on the Bluetooth packet error rate for single-slot packets that incorporates two inherent, short-term dependencies in the Bluetooth frequency hop sequence. First, the hop frequencies are chosen from a limited s...

We consider the reliability of a single-unit system whose cumulative damage over time is a continuous wear process that depends on an external environment process . We explicitly derive the failure time distribution and moments in terms of Laplace–Stieltjes transforms by analyzing the Markov additive process and demonstrate its applicability on an...

Modeling of activity duration has become an important aspect in characterizing activity and travel patterns. The standard approach for analyzing activity duration is to use hazard-based models to account for the duration dependence within an activity while estimating covariate effects and heterogeneity. The effects of covariates and heterogeneity o...

The paper presents a simple approximation technique for statistical tolerance analysis, namely, the allocation of component tolerances based on a known assembly tolerance. The technique utilizes a discretized, multivariate kernel density estimate and a simple transformation to approximate the probability distribution of the overall assembly charact...

Thesis (M.S.)--Ohio University, June, 1997.