Optimal Task Reallocation in Heterogeneous Distributed Computing Systems with Age-Dependent Delay Statistics

This paper describes a generalized variable neighborhood search heuristic for the Capacitated Vehicle Routing Problem with Stochastic Service Times, in which the service times at vertices are stochastic. The heuristic is tested on randomly generated instances and compared with two other heuristics and with an alternative solution strategy. Computational results show the superiority and effectiveness of the proposed heuristic.

Appointment scheduling with multiple providers and stochastic service times

Optimizing road network daily maintenance operations with stochastic service and travel times

We consider the problem of assigning job due-dates when the objective is to minimise the total expected earliness and tardiness penalty. This type of problem is important because modern production and distribution systems discourage both tardy and early deliveries. This paper extends a general due-date policy, the so-called SL rule, to incorporate not only tardiness but also earliness. We provide optimal policy parameters for the linear and quadratic penalty cases in an M/M/1 system.

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.

In a facility with front room and back room operations, it is useful to switch workers between the rooms in order to cope with changing customer demand. Assuming stochastic customer arrival and service times, we seek a policy for switching workers such that the expected customer waiting time is minimized while the expected back room staffing is sufficient to perform all work. Three novel constraint programming models and a shaving algorithm are presented. Experimental results show that the best constraint programming model, using shaving, is able to find and prove optimal solutions for almost all problem instances within a reasonable run-time, but that an existing heuristic algorithm performs better in terms of solution quality over time. A hybrid method combining the heuristic and the best constraint programming method is shown to perform better than either of these approaches separately. This is the first work of which we are aware that solves a queueing control problem with constraint programming.

This paper extends a state-based timing analysis for Synchronous Dataflow Applications on an MPSoC with shared memory. The existing approach transforms a mapped and timing annotated SDF graph into a timed automata representation for the analysis of timing properties. One major drawback of the existing timing annotation approach is the usage of best- and worst-case execution time intervals, resulting in an overestimation of the actual timing behavior. This paper proposes to replace the timing bound annotation with a probability density function. For the overall timing analysis we use a stochastic timed automata model. We demonstrate and evaluate our approach on a Sobel filter, which is used in many image and video processing algorithms. As a reference, we compare our stochastic execution time model against a fixed best-/worst-case execution time model and against the measured execution time on an FPGA prototype. The results are promising and clearly indicate that our probabilistic approach provides tighter timing analysis results in comparison to the best-/worst-case execution analysis model.

Efficient parallel simulation of a sliding window protocol

The National Institute of Standards and Technology (NIST) timescale produces a real-time realization of UTC(NIST) in the form of a pulse-per-second (PPS) time signal and a 5 MHz frequency reference. UTC(NIST) contributes to UTC (Coordinate Universal Time) using TWSTFT (Two-Way Satellite Time and Frequency Transfer) and common-view GPS (Global Positioning System) techniques. The NIST timing signal is also disseminated to the international and domestic communities using various other techniques: NIST Automated Computer Time Service (ACTS), Internet Time Service (ITS) and NIST Time Measurement and Analysis Service (TMAS). The accuracy of NIST contribution to UTC and of the dissemination of the NIST PPS time signal depends upon the quality of the calibration of the time delay between each user and the UTC(NIST) reference point. Hardware constraints, spatial constraints, and in some cases simple convenience, require that the physical input to the transfer and dissemination equipment be distributed throughout the NIST campus. A robust on-campus timing distribution and calibration system is required to ensure the accuracy and stability of these signals: for example, the current error budget for a typical time transfer calibration requires local delays to be known to an uncertainty of less than 200 ps. We utilize a clock trip calibration protocol to measure the delays introduced by the local time distribution system at NIST: a clock trip involves measuring the time difference between the UTC(NIST) reference point and a travelling clock, then transporting the travelling clock to a secondary reference point where another measurement of the time difference is performed. Clock trips are advantageous as they are a nonintrusive measurement of the full system delay. To achieve calibrations with an accuracy on the order 100 ps, we have developed a means to transport a cesium clock between remote locations on site with minimal vibration and an analysis protocol to maximize the information acquired by each clock trip. We will introduce the system used to distribute time signals at NIST, the hardware and protocol we have developed to calibrate secondary timing reference points, and data we have collected based on their implementation.

In this paper, we investigate the performance of batch service queue model with second optional service, repairable breakdown and warm standby server. Both primary operating and warm standby servers provide First Essential Service (FES) and Second Optional Service (SOS) to customers, wherein FES is all arriving customers and only some of them may further request the SOS. The service times, failure times and repair times of both primary operating and warm standby server are assumed to follow exponential distributions. We use Runge-Kutta method to obtain the transient state probabilities and matrix-decomposition method to obtain the steady-state probabilities of the model. Also, a cost model is presented to determine the optimal service rates so that the expected cost is minimised. Finally, the effect of the model parameters on the system behaviour is demonstrated through numerical results and discussions.

The analysis of a three-phase queueing servicing system with renewable servers

A node in a switch architecture is examined. The node receives messages, stores and processes them locally, and transmits them to other nodes. Both message transmission and processing times depend on the message length, which is a random variable. The node is modeled as a queue with a single server where the interarrival and service times are related, and different cases are studied where this relation ranges from direct dependence to moderate correlation. Exponential and uniform probability density functions (pdf's) are examined, and the results are compared to the independent model. A relationship is developed for the system time pdf of a message, which is used to find the system time pdf and the system state probabilities. These quantities are directly related to buffer sizes and transmission delays within the architecture. >

Utilities are increasingly considering the use of temporary water transfers to augment their supplies during periods of drought, an alternative that is often less expensive than expanding safe yields through new infrastructure. Understanding the volume and timing of transfers is important for developing contracts between buyer and seller and can be challenging due to the transient nature of drought, a situation complicated by the uncertainties associated with climate change. While transfer arrangements have received some attention in the literature, the effects of climate change on such agreements remain unexplored. This paper investigates these impacts using an improved method for developing new hydro-climate scenarios. A technique for producing stochastic time series of inflows is described, one which effectively replicates the autocorrelation present in the historic record. Unlike autoregressive (and similar) models that assume complete stationarity, the modified fractional Gaussian noise (mFGN) method preserves the seasonal patterns in the correlation structure, thereby providing some advantages when modifying historical streamflow records to reflect alternative hydro-climate scenarios. Alternative scenarios are developed for the Research Triangle region of North Carolina, an area with several utilities currently seeking to use a system of risk-based transfer agreements as a means to meet demand during droughts. This study simulates hydrologic conditions and transfer activity from 2010 to 2025 under a variety of climate scenarios. Results indicate that increased variability in inflows, with no change in the mean, corresponds to slight increases in transfer activity. However, when increased variability is paired with modest decreases in expected inflows (7%), transfer activity is doubled.

We introduce a new flexible statistical model for disability and reliability case studies, designed to enhance modeling capabilities in reliability engineering and disability prevalence analysis. This model extends the classical Pareto Type Ⅱ distribution by incorporating additional shape parameters, allowing for greater flexibility in modeling various hazard rate shapes and tail behaviors, particularly light tails. Essential properties of the model are explored. The model's applicability and risk assessment potential are demonstrated through case studies: one focusing on reliability engineering (aircraft windshield failure and service times) and another analyzing disability prevalence rates in Saudi Arabia for 2016. Advanced tail analysis tools such as the Hill estimator, Value-at-Risk (VaR), and Peaks Over a Random Threshold (PORT) were employed. The analyses reveal that the proposed OGPTII model offers superior goodness-of-fit compared to several existing models for both reliability and Saudi Arabia disability data, accurately captures the light-tailed nature of the datasets, and provides crucial insights into potential risk factors. This makes it a valuable tool for predictive maintenance strategies in engineering and for evidence-based policy planning and resource allocation in the context of disability management.

In this paper a two-station, single-server queuing system with random breakdowns is investigated. Two classes of customers arrive at the system with finite buffers where one server is dedicated to only one customer class while the second server can process both types of customers. Interarrival and service times are exponentially distributed while failure and repair times follow the generalized exponential distribution. Motivated by the application of alternative tools in semiconductor manufacturing two different control schemes are considered that trigger processing of customers of the second class on the server that is usually dedicated to customers of the first class. The first two moments of customer flow times for both control schemes are calculated.