Skip to content
Free planNo credit card requiredStart writing
AI for Research home
AI for Research home
Dashboard
Sign inGet started

Search 250 million research papers

Read the abstract, see who wrote it and who cites it, and get the citation in APA, MLA, Chicago or Harvard.

About 5,270 papers for “Processor scheduling”

Processor scheduling

Rosti, E., Serazzi, G, Consiglio Nazionale delle Ricerche, Rome (Italy). Progetto Finalizzato: Sistemi Informatici e Calcolo Parallelo · OpenGrey (Institut de l'Information Scientifique et Technique) · 1994

Cited by 0

Eclipse Processor Scheduling

Martijn J. Rutten, Jos T. J. Van Eijndhoven, Evert-jan D. Pol · 2008

Abstract—Eclipse is an architectural framework for on-chip CPU and coprocessor communication, combining application configuration flexibility with the efficiency of functionspecific hardware. A Kahn process network application model is supported by a generic communication infrastructure and function-specific processors with multi-tasking capabilities. High data bandwidth and limited stream buffer memory lead to high buffer synchronization rates and high task switch rates. These are supported by shells dedicated to each processor. The shells each implement a task scheduler for a dynamic workload environment with guarantees for minimum resource budgets, and achieve effective run-time task selection in a few clock cycles. The resulting implementation with distributed shells provides good scalability. Index Terms—Dataflow, synchronization, multi-tasking, non-preemptive scheduling, justice. I.

Cited by 0

Shared processor scheduling

Dereniowski, Dariusz, Kubiak, Wieslaw · arXiv (Cornell University) · 2017

We study the shared processor scheduling problem with a single shared processor where a unit time saving (weight) obtained by processing a job on the shared processor depends on the job. A polynomial-time optimization algorithm has been given for the problem with equal weights in the literature. This paper extends that result by showing an $O(n \log n)$ optimization algorithm for a class of instances in which non-decreasing order of jobs with respect to processing times provides a non-increasing order with respect to weights --- this instance generalizes the unweighted case of the problem. This algorithm also leads to a $\frac{1}{2}$-approximation algorithm for the general weighted problem. The complexity of the weighted problem remains open.

Cited by 0Open access

Shared processor scheduling

Dariusz Dereniowski, Wiesław Kubiak · Journal of Scheduling · 2018

We study the shared processor scheduling problem with a single shared processor to maximize total weighted overlap, where an overlap for a job is the amount of time it is processed on its private and shared processor in parallel. A polynomial-time optimization algorithm has been given for the problem with equal weights in the literature. This paper extends that result by showing an $$O(n \log n)$$ -time optimization algorithm for a class of instances in which non-decreasing order of jobs with respect to processing times provides a non-increasing order with respect to weights—this instance generalizes the unweighted case of the problem. This algorithm also leads to a $$\frac{1}{2}$$ -approximation algorithm for the general weighted problem. The complexity of the weighted problem remains open.

Cited by 3Open access

Video signal processor scheduling

Kock, EA Erwin de, Aarts, EHL Emile, Verstraten, AJE Angelica · TU/e Research Portal (Eindhoven University of Technology) · 2001

Cited by 0

Processor Allocation, Message Scheduling, and

Algorithm Selection For, Jack M. West · 2007

VI TABLES ...............................................................................................................................VIII FIGURES .................................................................................................................................X CHAPTER I ..............................................................................................................................1

Cited by 0

Processor Allocation and Scheduling in

Distributed Systems Mohankumar, Mohankumar Zade, Nihal Ch et al. · 2007

this paper. This section also comments on the difference between the terms processor allocation and scheduling in context of distribute systems. Section III attempts to categorize the algorithms proposed in various categories for better understanding, comparison and future works. In section IV some example algorithms are discussed and section V briefly describes some example processor allocation algorithms of various classes. section VI presents co-scheduling algorithm. Lastly, summary and conclusions are presented in section VII

Cited by 0

Scheduling with Limited Processor Availability

n.n. · 2013

In scheduling theory the basic model assumes that all machines are continuously available for processing throughout the planning horizon. This assumption might be justified in some cases but it does not apply if certain maintenance requirements, breakdowns or other constraints that cause the machines not to be available for processing have to be considered. In this chapter we discuss results related to deterministic scheduling problems where machines are not continuously available for processing. Examples of such constraints can be found in many areas. Limited availabilities of machines may result from pre-schedules which exist mainly because most of the real world resources planning problems are dynamic. A natural approach to cope with a dynamic environment is to trigger a new planning horizon when the changes in the data justify it. However, due to many necessities, as process preparation for instance, it is mandatory to take results of earlier plans as fixed which obviously limits availability of resources for any subsequent plan. Consider e.g. ERP (Enterprise Resource Planning) production planning systems

Cited by 0

Task scheduling in parallel processor systems

Syarifah Zyurina Nordin · eSpace (Curtin University) · 2011

Task scheduling in parallel processing systems is one of the most challenging industrial problems. This problem typically arises in the manufacturing and service industries. The task scheduling problem is to determine a set of task assignments to a set of parallel processors for execution so as to optimize a specified performance measure. The difficulty of the problem is that the scheduling needs to satisfy a set of requirements as well as a range of environmental constraints. The problem is known to be NP-complete.In this study, we consider a non-preemtive task scheduling problem on iden- tical and unrelated parallel processor systems. We are interested in the objective function that minimizes the maximum of the completion time of the entire set of tasks (i.e makespan) so as to ensure a good load balance on the parallel proces- sors. We consider three different task characteristics to the classical task scheduling problem that has a set of n independent tasks to be assigned to m parallel processors.The first task characteristic that we consider is an on-line scheduling with release date specifications on an identical parallel processing system with a centralized queue and no splitting structure. We focus on developing simple and efficient heuristics for this problem. Three heuristic algorithms are proposed to solve this non-deterministic problem with scheduling over time where the availability of each task is restricted by release date. Our approach uses a multi-step method in the task selection phase and a greedy search algorithm in the processor selection phase. The multi-step method is used to reduce the non-determinism in on-line scheduling by partitioning the scheduling process into several procedures. We introduce two procedures in the priority rule loop which we refer to as Cluster Insertion and Local Cluster Interchange.Computational testing on randomly generated data is conducted using Microsoft Visual C++ 6.0 to examine the effectiveness of the proposed multi-step method against the optimal solution. Different size of problems are tested in the experiment involving 3 processors by 200 tasks up to 5 processors by 1000 tasks with five clusters ranging from 10 to 50. The computational results show that all the three heuristics performed very well with the value of the average gaps are improved as the number of the tasks in the system is increases. The average gap for all the three heuristics are less than 1.04% for the largest tested cases (i.e for 1000 tasks run on 5 processors).In the second problem, we address priority consideration as an added feature to the basic task characteristics of unrelated parallel processors scheduling. The priority consideration is defined by a list of ordered independent tasks with priority. A task requires to start processing after another task is finished on the same processor based on priority but may require to start earlier if processed on other machine. Our aim is to develop Mixed Integer Linear Programming models to obtain optimal solutions for three type of priority lists which are ascending order, descending order and general priority list. We validate the model using a case study taken from the literature. Then, computational testing is implemented on the general priority list using AIMMS 3.10 package and CPLEX 12.1 as the solver. Computational results show that the proposed MILP model is effective and produces optimal results for all tested cases. The model is very efficient as 95% of all the instances, which are problem up to 80 tasks assigned on 5 processors, have been solved within 5 minutes of CPU time.In the final problem, we address a further problem for the task scheduling with a disruption problem that occurs on the parallel processor system. The disruption is causes by the unavailability of the processor during a certain time and it is called resource disruption. Our recovery solution for the disruption problem is a rescheduling approach. A MILP model is developed for the rescheduling model for the case of non-resumable tasks. Recovery model for the disrupted initial schedule with dummy insertion is proposed for predictive disruption management and match up schedule for post-disruption management. To evaluate the model, computational testing is performed with different sets of data.Different levels of disruptions are considered with different weights in the objective function to observe the stability of the model. The optimum initial schedule and the rescheduling model is performed using CPLEX 12.1 solver in AIMMS 3.10 package. In our computational results we measure the stability rate which is to observe the stability condition of the current schedule compared to the initial schedule in term of the task migration. From the results, the stability is improved when the number of tasks in the system increases within a reasonable amount of time. Another interesting observation is that our model yields small average gaps that are less than 7.99% within 300 seconds of the CPU time for a large data set that reach 200 tasks by 10 processors. The average gaps are considerably small for the disruption problem since the rescheduling model have to match up with the optimum initial schedule.

Cited by 0Open access

A genetic approach for processor scheduling

Tabassum A. Maktum, Rashmi A. Dhumal, Lata Ragha · 2014

CPU scheduling is one of the most important tasks of operating system, which provides a specific sequence of execution for the processes waiting in queue. The scheduling of processes has very much impact on efficiency and performance of CPU. The major objective of this task is to maximize the efficiency and hence the performance. We present a Genetic Algorithm based approach for CPU scheduling. Genetic algorithm is evolutionary technique which finds optimal solutions for NP hard problems. Genetic Algorithm based approach presented here, is used to find near optimal solution for the scheduling problem. The algorithm schedules the processes based on their average waiting time. We implement and compare the results of two traditional algorithms: FCFS (First Come First Serve) and SJF (Shortest Job First) with Genetic algorithm based CPU scheduling.

Cited by 7

Processor scheduling in shared memory multiprocessors

John Zahorjan, Cathy McCann · 1990

Existing work indicates that the commonly used “single queue of runnable tasks” approach to scheduling shared memory multiprocessors can perform very poorly in a multiprogrammed parallel processing environment. A more promising approach is the class of “two-level schedulers” in which the operating system deals solely with allocating processors to jobs while the individual jobs themselves perform task dispatching on those processors.

Cited by 154Open access