Task Scheduling Research Articles

Integrated scheduling of multi-site repair tasks considering customer preferences under a novel two-dimensional combination warranty strategy

Current warranty strategies typically focus on the interests of either the manufacturer or the customer, and fail to meet the diverse needs of the customers, as they only consider a single reimbursement or repair method. Furthermore, these strategies neglect out-of-warranty customers, resulting in a loss of profit for manufacturers. To serve both in-warranty and out-of-warranty customers, a novel two-dimensional combination warranty strategy is proposed in this paper which combines two reimbursement methods and two repair methods. Under the two-dimensional combination warranty strategy, we establish an integrated scheduling model for determining repair strategies and allocating service resources, referred to as 2DCW-MSRASM. In 2DCW-MSRASM, the manufacturer arranges appropriate warranty strategies for in-warranty orders according to the corresponding product conditions (age and usage), while for out-of-warranty customers, the manufacturer makes decisions based on its own resource situation and customer preferences. The objective of 2DCW-MSRASM is to achieve an optimal scheduling that maximizes the manufacturer’s profit and the satisfaction of both in-warranty and out-of-warranty customers. In this model, TODIM (an acronym in Portuguese of Interactive and Multi-Criteria Decision Making) is utilized to evaluate the repair services from the customer’s perspective, which considers the limited rationality of decision-makers. An improved non-dominated neighbor immune algorithm (INNIA) is developed to solve the model, incorporating an initialization method and three problem-oriented local search operators to enhance search capability. Following that, extensive experiments are conducted to demonstrate the superiority of INNIA. Eventually, the proposed strategy is demonstrated to possess noteworthy advantages when the order size is large and the proportion of in-warranty orders is high.

Optimizing low-power task scheduling for multiple users and servers in mobile edge computing by the MUMS framework

In today's increasingly popular Internet of Things (IoT) technology, its energy consumption issue is also becoming increasingly prominent. Currently, the application of Mobile Edge Computing (MEC) in IoT is becoming increasingly important, and scheduling its tasks to save energy is imperative. To address the aforementioned issues, we propose a Multi-User Multi-Server (MUMS) scheduling framework aimed at reducing the energy consumption in MEC. The framework starts with a model definition phase, detailing multi-user multi-server systems through four fundamental models: communication, offloading, energy, and delay. Then, these models are integrated to construct an energy consumption optimization model for MUMS. The final step involves utilizing the proposed L1_PSO (an enhanced version of the standard particle swarm optimization algorithm) to solve the optimization problem. Experimental results demonstrate that, compared to typical scheduling algorithms, the MUMS framework is both reasonable and feasible. Notably, the L1_PSO algorithm reduces energy consumption by 4.6 % compared to Random Assignment and by 2.3 % compared to the conventional Particle Swarm Optimization algorithm.

A deep reinforcement learning approach for dynamic task scheduling of flight tests

A deep reinforcement learning approach for dynamic task scheduling of flight tests

Research on Task Scheduling for Internet of Things Cloud Computing Based on Improved Chicken Swarm Optimization Algorithm

Aiming at the shortcomings of long completion time and high consumption cost of cloud computing batch task scheduling in IoT, an Improved Chicken Swarm Optimization Algorithm (ICSO) for task scheduling in cloud computing scenarios is proposed. Specifically, in order to solve the problems of slow convergence and falling into local optimum of the chicken swarm optimization algorithm, we adopt the nonlinear decreasing technique of the rooster and the weighting technique of the hen, optimize the following coefficients of the chicks, and apply ICSO to cloud computing task scheduling. In simulation experiments, we conducted a large number of experiments using four standard benchmark functions with different number of tasks and the results show that ICSO algorithm reduces 25.8%, 9.3%, 8.8%, 7.5% in small task time compared to CSO, DCSO, GCSO, ABCSO in large task time by 30.8%, 8.3%, 7.8%, 6.3%, 11.8%, 10.3%, 8.8%, 7.5% savings in small task cost and 25.8%, 11.2%, 10.8%, 9.3% savings in large task cost. This method effectively reduces task scheduling time and cost consumption. Meanwhile, we tested it in combination with an IoT-based cloud platform and achieved very satisfying Results.

Scalability and Performance Optimization Strategies in Large-Scale System Architectures

This paper aims to break away from traditional thinking paradigms by introducing theories and methods from complexity science to provide innovative and effective solutions for scalability and performance optimization of large-scale system architectures. The paper first constructs a complex network model for large-scale system architecture, which includes nodes (services), edges (interactions), and global attributes, along with dynamic models reflecting key metrics such as system performance and scalability, providing a theoretical foundation for subsequent research. Subsequently, from a complexity perspective, it innovatively proposes scalability strategies, including adaptive scalability strategies for complex network optimization, data-driven complex system regulation, and elastic computing and service orchestration mechanisms. In terms of performance optimization, it applies tools such as complex systems observation theory and network flow theory to explore insights and breakthroughs in complexity related to full-link performance monitoring and diagnostics, asynchronous communication, task scheduling, and software-hardware collaborative optimization. Finally, the paper summarizes the research findings, distilling the core concepts and methodological innovations of scalability and performance optimization strategies from a complexity perspective. It also looks ahead to the prospects of further integrating complexity science and information technology in the future, along with potential research topics in new scenarios and challenges.

Enhanced Whale Optimization Algorithm for task scheduling in cloud computing environments

The escalation of cloud services, driven by their accessibility, improved performance, and cost-effectiveness, has led cloud service providers to consistently seek methods to expedite job completion, thereby boosting profits and reducing energy consumption expenses. Despite developing numerous scheduling algorithms, many of these techniques address only a specific objective within the scheduling process. To efficiently achieve better optimization results for the cloud task scheduling problem, a novel approach, the Enhanced Whale Optimization Algorithm (EWOA), is proposed. EWOA integrates the WOA with the Lévy flight. The incorporation of Lévy flight is tailored to broaden the search space of WOA, expediting convergence with adaptive crossover. The EWOA model is simulated using the Cloudsim tool and evaluated under diverse test conditions. The effectiveness of EWOA is assessed by employing various parameters and comparing them with existing algorithms. The results demonstrate that EWOA outperforms other algorithms in resource utilization, energy consumption, and execution cost, establishing its superiority in addressing the complexities of multi-objective cloud task scheduling.

Automated Data Quality Control System in Health and Demographic Surveillance System

The automated data quality management system serves as a comprehensive solution developed to enhance the precision, dependability, and uniformity of data within data-driven organizations. Such systems play an important role in eliminating the shortcomings associated with manual data quality management, which is prevalent in health and demographic surveillance systems (HDSS). The ongoing difficulty of ensuring data quality through manual processing hinders the HDSS's capacity to optimize data quality effectively. To address this challenge, our study adopted design science methodologies to provide guidelines for the design and implementation of the automated data quality control system. The open source technologies (Pentaho data integration, R Studio, SQL, Windows task scheduler) were used to facilitate the automation and validation of the incoming and database resident data. The quality of data has vastly improved since the implementation of the proposed system. The findings suggest that the automated data quality control system exhibited superior performance compared to the manual methods, thereby minimizing errors and time-wasting efforts.

An integrated optimization method to task scheduling and VM placement for green datacenters

In the realm of cloud computing, effective resource allocation can significantly enhance the energy efficiency of datacenters. Task scheduling and Virtual Machine Placement (VMP) are two pivotal aspects of resource allocation. However, in current research, they are often treated separately, overlooking the potential for integrated optimization. In this paper, we propose an integrated solution for task scheduling and VMP in energy-efficient datacenters, based on queueing theory and Deep Reinforcement Learning (DRL) methods. This novel and comprehensive approach provides an alternative perspective for resource scheduling strategies in datacenters. We construct a queueing theory model for task scheduling, aiming to minimize the number of VMs that need to be instantiated, while ensuring that Service Level Agreement (SLA) violation remains at a low level. Furthermore, we design a VMP algorithm based on DRL for real-time selection of Physical Hosts (PHs) for deploying VMs. Finally, we conduct a simulation evaluation using a small-scale datacenter. The experimental results demonstrate that our method consistently ensures a lower rate of SLA violation. Compared to existing algorithms, the DRL-based VMP algorithm enables a more balanced utilization of the various resources in the PHs and reduces the total power consumption of the datacenter by more than 10% on average.

Joint multi-server cache sharing and delay-aware task scheduling for edge-cloud collaborative computing in intelligent manufacturing

Joint multi-server cache sharing and delay-aware task scheduling for edge-cloud collaborative computing in intelligent manufacturing

Enhanced Machine Learning for Cloud Computing Electricity Cost Prediction

Cloud computing is becoming popular in the tech industry as it offers convenient computing without the need to buy physical hardware. Instead, companies provide these services using their own computers and servers powered by electricity. However, running these data centers consumes a lot of energy, and with rising electricity prices, minimizing energy usage has become a big challenge. One way to tackle this is by efficiently managing data storage and scheduling tasks. In this article, we suggest using an advanced machine learning model called Extreme Gradient Boosting (XGBoost) to predict electricity prices and optimize data storage, helping to reduce energy costs in data centers

Dynamic edge clustering and task scheduling for edge assisted metaverse system in the field of remote work and collaboration

SummaryThe metaverse, a future digital world for living, working, learning, and interacting, is rapidly gaining significance, particularly in the domain of remote work and collaboration. This emerging digital landscape demands high‐performance, low‐latency, and scalable services to provide an immersive user experience. The current technology used in metaverse systems has some limitations, which emphasize the importance of adopting emerging technologies like Edge Computing (EC). However, as the number of users and data volume increases, it can impact both system performance and scalability of the edge‐assisted metaverse system. Additionally, the uneven distribution of edge servers can cause inconsistencies and result in high latency. To overcome these challenges, this paper proposes a dynamic edge clustering and task scheduling approach for edge‐assisted metaverse systems (DTAM) in the field of remote work and collaboration. The proposed approach addresses the challenges of high user volume and uneven resource distribution by incorporating dynamic clustering and edge server assistance to improve clustering performance. Furthermore, a Prioritized Experience Replay‐based Deep Q‐learning algorithm with state augmentation (PERDQSA) for task scheduling is introduced to improve sample efficiency and performance. The performance of the proposed DTAM is evaluated against existing techniques, and experimental results demonstrate its significant superiority in terms of specific metrics such as bandwidth, task response time, energy efficiency, and latency. The experiments demonstrated that DTAM outperforms Transformation‐based Edge Computing Deep Q‐Learning (TransEC‐DQL) in several key metrics. Specifically, DTAM achieves 28.5% reduction in latency, 13.7% reduction in response time, and 6.4% improvement in bandwidth compared to TransEC‐DQL. These results signify that DTAM can deliver a significantly enhanced user experience in the metaverse, particularly in the context of remote work and collaboration.

A simulation-based genetic algorithm for a semi-automated warehouse scheduling problem with processing time variability

For warehouse operations, efficiently scheduling the available resources is crucial to improve the productivity and customer satisfaction. This paper proposes a simulation-based evolutionary algorithm for order scheduling and multi-robot task assignment in a robotic mobile fulfillment system. The algorithm proactively deals with the effects of the processing time variability by evaluating schedules based on both its system performance as well as its robustness under uncertain conditions. The algorithm implements an efficient resource allocation method and a variance reduction technique to reduce the overall computational burden. The experimental results show that the techniques to reduce the computational time are effective and can significantly reduce the amount of simulations required for the fitness evaluation. If a candidate schedule is allocated insufficient simulation replications it can lead to an inaccurate estimate of its long-term average performance. This could lead to an average performance loss of 7.3 %. Furthermore, the proactive scheduler is able to generate schedules that are more robust compared to deterministically generated. A reduction in the average operational cost of about 5 % can be reached, compared to a deterministically generated schedule. The paper reveals the relevance of identifying and modeling uncertainty when designing schedules in an operational system, rather than looking for optimal schedules for ideal scenarios.

Model and heuristics for the multi-manned assembly line worker integration and balancing problem

This paper examines the balancing of assembly lines with multi-manned stations and a heterogeneous workforce. Both topics received considerable attention in the literature, but not in an integrated fashion. Combining these two characteristics gives rise to a highly combinatorial Multi-manned Assembly Line Worker Integration and Balancing Problem. When considering multi-manned stations, the already coupled decisions on assigning tasks to heterogeneous workers and workers to stations must be further linked with task scheduling assessments. We propose a Mixed-Integer Linear Programming model and develop two heuristic solution procedures, which tackle the problem with a hierarchical decomposition approach. Computational tests on a large dataset indicate that the proposed method can obtain good primal bounds in short computational times. We demonstrate that these results can be applied to the monolithic model either as a warm start or in a proximity search procedure to obtain synergistic gains with statistically significant differences. From a managerial perspective, we show that multi-manned stations can reduce the assembly line's length even in the presence of a heterogeneous workforce, which is crucial for many industries manufacturing large-size products.

Optimal design and scheduling of ecofriendly process for the production of natural tocopherols present in residue from the vegetable oil refinery

This paper addresses the optimal design of molecular distillation (MD) process for the treatment of a residue of the edible oils refining industry. This work focuses on optimizing the number of units and their size, performing the scheduling of the separation tasks assigned to each unit with the objective of treating a residue with the minimum number of process equipment at a specific size to process a certain annual production of ODD (Oil Deodorization Distillate). The MD is considering clean technology because only consumption electric energy and not generate highly polluting waste. The stream of ODD has a high organic contaminant load. This requires more than one distillation stage to obtain tocopherols as a product (four stage for this study case). Here, to evaluate the process with minimum number stage a first screening was done using the traditional model of constant size factors, but afterwards the best candidate alternative was optimized as regards to the flow rates to be recycled to previous stages. The objective is to add value to a residual stream of oil refining processes, resorting to MD technology, given that MD is a very expensive technology, and in this particular case where its size is small, limited by manufacturer. For all scheduling alternatives analyzed, the best economic performance yields a Profit of 381,360 $/yr with two MD equipment and optimizing the material balances and recycles. This process is with an assignment of tasks to units that yields one larger equipment to perform the first three separation stages, and a quite small equipment to perform the last separation stage. This process has even better economic performance than the same process without reflux recirculation optimization. The process with one unit yielded technically infallible results. The process with three units showed less favorable economic performance.

A hybrid optimization algorithm for energy-aware multi-objective task scheduling in heterogeneous multiprocessor systems

A hybrid optimization algorithm for energy-aware multi-objective task scheduling in heterogeneous multiprocessor systems

Task Scheduling in a Cloud Environment: A comparative Study

In this study, we explore the application of Monarch Butterfly Optimization (MBO) algorithms for task scheduling in cloud computing, comparing its performance against widely used optimization techniques, namely Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO).Task scheduling in the cloud is a critical aspect influencing resource utilization, turnaround time, and overall system efficiency. MBO, known for its effective exploration- exploitation balance, is examined for its suitability in addressing the complexities of cloud computing environments. The study investigates MBO's advantages, such as enhanced adaptability to dynamic conditions, effective handling of multi-objective optimization, and its consideration of bandwidth as a critical resource. Comparative analyses with ACO and PSO highlight MBO's superior performance in achieving near-optimal task schedules, emphasizing its potential to offer innovative solutions to the challenges posed by task scheduling in dynamic and resource-constrained cloud environments. This research contributes valuable insights into the strengths of MBO, paving the way for advancements in optimization methodologies tailored for cloud computing systems.

Qsmix: Q-learning-based task scheduling approach for mixed-critical applications on heterogeneous multi-cores

Qsmix: Q-learning-based task scheduling approach for mixed-critical applications on heterogeneous multi-cores

Contextual Information Based Scheduling for Service Migration in Mobile Edge Computing

Mobile Edge Computing (MEC) is a distributed computing paradigm that delivers processing and data storage capabilities closer to the network edge, which is adjacent to mobile consumers and devices. MEC lowers latency, reduces data transmission times, and improves overall performance for mobile apps by relocating computing resources to the network’s edge. But, due to higher average load and longer elapsed time, modern end devices such as smartphones and tablets cause major load challenges in mobile computing networks. Furthermore, if smartphones cause unpredictable traffic patterns, it becomes impossible to model and forecast the nature of communication. Such confusing traffic figures are caused not just by bursty Internet traffic, but also by multitasking operating systems that allow users to swiftly switch between active apps. Mobility of users and end devices impose a difficult challenge to provide continuous services in mobile computing. In this paper, this issue is addressed using the Contextual Information Based Scheduling (CIBS) technique to optimally allocate resources and provide seamless service to the users. The proposed method is implemented with NS-3, an open-source network simulator that provides a comprehensive set of modules for Mobile Edge Computing (MEC) simulations, including mobility modelling support. The experimental results show that CIBS offers migration time of 97512ms, delay time of 372115ms, execution time of 1061328ms and downtime of 98715ms. The results are compared with the existing Mobility-Aware Joint Task Scheduling (MATS) approach. The obtained results show that CIBS outperforms MATS with regard to migration time, latency, execution time and downtime.

Integrated decision of production scheduling and condition-based maintenance planning for multi-unit systems with variable replacement thresholds

In practice, production scheduling and maintenance planning are two popular technologies that are interdependent, but they are most often performed independently. In this study, we have developed an integrated model for coordinating single-machine scheduling and condition-based maintenance planning. Firstly, we build upon existing scheduling research that often views the system as a whole and further consider the degradation and maintenance of multiple units within the system during the scheduling process. Secondly, in contrast to traditional degradation modeling which usually assumes a single threshold to trigger maintenance/replacement, we also consider the variability of replacement thresholds caused by the heterogeneity of scheduling tasks. Based on this foundation, we propose an integrated policy that considers both scheduling and condition-based maintenance. Furthermore, we have developed an integrated decision model for multi-unit systems. The model aims to minimize the total weighted expected completion time over the entire scheduling time horizon. Moreover, the genetic algorithm was employed to optimize the integrated model. Finally, we present an example of an SA adjustable multi-axis machine and conduct a sensitivity analysis of key parameters to validate the integration policy.

Comparative analysis of real issues in open-source machine learning projects

ContextIn the last decade of data-driven decision-making, Machine Learning (ML) systems reign supreme. Because of the different characteristics between ML and traditional Software Engineering systems, we do not know to what extent the issue-reporting needs are different, and to what extent these differences impact the issue resolution process.ObjectiveWe aim to compare the differences between ML and non-ML issues in open-source applied AI projects in terms of resolution time and size of fix. This research aims to enhance the predictability of maintenance tasks by providing valuable insights for issue reporting and task scheduling activities.MethodWe collect issue reports from Github repositories of open-source ML projects using an automatic approach, filter them using ML keywords and libraries, manually categorize them using an adapted deep learning bug taxonomy, and compare resolution time and fix size for ML and non-ML issues in a controlled sample.Result147 ML issues and 147 non-ML issues are collected for analysis. We found that ML issues take more time to resolve than non-ML issues, the median difference is 14 days. There is no significant difference in terms of size of fix between ML and non-ML issues. No significant differences are found between different ML issue categories in terms of resolution time and size of fix.ConclusionOur study provided evidence that the life cycle for ML issues is stretched, and thus further work is required to identify the reason. The results also highlighted the need for future work to design custom tooling to support faster resolution of ML issues.