The Experts below are selected from a list of 360 Experts worldwide ranked by ideXlab platform

Albert Y. Zomaya - One of the best experts on this subject based on the ideXlab platform.

  • towards decentralized Load Balancing in a computational grid environment
    Lecture Notes in Computer Science, 2006
    Co-Authors: Kai Lu, Riky Subrata, Albert Y. Zomaya
    Abstract:

    Load Balancing has been a key concern for locally distributed multiprocessor systems. The emergence of computational grid extends this Problem, such as scalability, heterogeneity of computing resources and considerable communication delay. In this paper, we study the Problem of scheduling a large number of CPU-intensive jobs on such systems. The time spent by a job in the system is considered as the main issue that needs to be minimized. The proposed dynamic algorithm of scheduling jobs consists of two policies: Instantaneous Distribution Policy (IDP) and Load Adjustment Policy (LAP). Our algorithm does not address directly the Load Balancing Problem since it is completely unrealistic in such large environments, but we will show that even a non-perfectly Load balanced system can behave reasonably well by taking into account the jobs' time demands. The proposed algorithm is evaluated by a series of simulations.

  • Observations on using genetic algorithms for dynamic Load-Balancing
    IEEE Transactions on Parallel and Distributed Systems, 2001
    Co-Authors: Albert Y. Zomaya, Yee Hwei Teh
    Abstract:

    Load-Balancing Problems arise in many applications, but, most importantly, they play a special role in the operation of parallel and distributed computing systems. Load-Balancing deals with partitioning a program into smaller tasks that can be executed concurrently and mapping each of these tasks to a computational resource such a processor (e.g., in a multiprocessor system) or a computer (e.g., in a computer network). By developing strategies that can map these tasks to processors in a way that balances out the Load, the total processing time will be reduced with improved processor utilization. Most of the research on Load-Balancing focused on static scenarios that, in most of the cases, employ heuristic methods. However, genetic algorithms have gained immense popularity over the last few years as a robust and easily adaptable search technique. The work proposed here investigates how a genetic algorithm can be employed to solve the dynamic Load-Balancing Problem. A dynamic Load-Balancing algorithm is developed whereby optimal or near-optimal task allocations can "evolve" during the operation of the parallel computing system. The algorithm considers other Load-Balancing issues such as threshold policies, information exchange criteria, and interprocessor communication. The effects of these and other issues on the success of the genetic-based Load-Balancing algorithm as compared with the first-fit heuristic are outlined.

Karen Dragon Devine - One of the best experts on this subject based on the ideXlab platform.

  • new challanges in dynamic Load Balancing
    Applied Numerical Mathematics, 2005
    Co-Authors: Karen Dragon Devine, Bruce Hendrickson, Erik G Boman, Robert Heaphy, James D Teresco, Jamal Faik, Joseph E Flaherty, Luis G Gervasio
    Abstract:

    Data partitioning and Load Balancing are important components of parallel computations. Many different partitioning strategies have been developed, with great effectiveness in parallel applications. But the Load-Balancing Problem is not yet solved completely; new applications and architectures require new partitioning features. Existing algorithms must be enhanced to support more complex applications. New models are needed for non-square, non-symmetric, and highly connected systems arising from applications in biology, circuits, and materials simulations. Increased use of heterogeneous computing architectures requires partitioners that account for non-uniform computing, network, and memory resources. And, for greatest impact, these new capabilities must be delivered in toolkits that are robust, easy-to-use, and applicable to a wide range of applications. In this paper, we discuss our approaches to addressing these issues within the Zoltan Parallel Data Services toolkit.

  • dynamic Load Balancing in computational mechanics
    Computer Methods in Applied Mechanics and Engineering, 2000
    Co-Authors: Bruce Hendrickson, Karen Dragon Devine
    Abstract:

    In many important computational mechanics applications, the computation adapts dynamically during the simulation. Examples include adaptive mesh refinement, particle simulations and transient dynamics calculations. When running these kinds of simulations on a parallel computer, the work must be assigned to processors in a dynamic fashion to keep the computational Load balanced. A number of approaches have been proposed for this dynamic Load Balancing Problem. This paper reviews the major classes of algorithms and discusses their relative merits on Problems from computational mechanics. Shortcomings in the state-of-the-art are identified and suggestions are made for future research directions.

Saoussen Krichen - One of the best experts on this subject based on the ideXlab platform.

  • solving a Load Balancing Problem with a multi objective particle swarm optimisation approach application to aircraft cargo transportation
    International Journal of Operational Research, 2016
    Co-Authors: Nadia Dahmani, Saoussen Krichen
    Abstract:

    To make air cargo carrier operations cost-effective, a challenging task consists on making profitable the stowage step of the transported freight. This can be accomplished by maximising the weight of the Loaded cargo. Owing to the nature and the urgency of the carried cargo, a priority level can also constitute a relevant potential objective to be maximised. In this paper, we present a two level Load Balancing Problem that consists in Loading items into containers which are then stowed in cargo holds of an aircraft. Two objectives are maximised: the total weight and the total priority of Loaded cargo. In order to minimise fuel consumption and satisfy stability requirements, a Load Balancing constraints are expressed in terms of the deviation between the gravity centre after Loading and its ideal position. An integer linear programming-based formulation is presented for the Problem at hand. The Loading process performs a discrete multi-objective particle swarm optimisation approach. In order to show the effectiveness of our algorithm and due to the importance of satisfying the Load Balancing constraints, a practical case study is addressed. An experimental investigation of our approach shows that the proposed approach performs well.

  • a particle swarm optimization approach for the bi objective Load Balancing Problem
    Electronic Notes in Discrete Mathematics, 2010
    Co-Authors: Saoussen Krichen, Nadia Dahmani
    Abstract:

    Abstract We propose in this paper a two level Loading Problem that consists in packing items into containers, then stowing these containers in an aircraft while maximizing both of the total weight and the priority of the Loaded cargo. At the same time, the center of gravity should be within a reasonable distance from the balance ideal position. We state the mathematical formulation of the Problem. The minimization of the number of containers is tackled using a multi-objective placement heuristic and the Loading process performs a discrete multi-objective particle swarm optimization approach. An experimental investigation is performed on various test instances to illustrate the effectiveness of our algorithm in solving the bi-objective Loading Problem.

Jesper Larsson Traff - One of the best experts on this subject based on the ideXlab platform.

  • what mpi could and cannot do for mesh partitioning on non homogeneous networks
    Lecture Notes in Computer Science, 2006
    Co-Authors: Guntram Berti, Jesper Larsson Traff
    Abstract:

    We discuss the mesh-partitioning Load-Balancing Problem for non-homogeneous communication systems, and investigate whether the MPI process topology functionality can aid in solving the Problem. An example kernel shows that specific communication patterns can benefit substantially from a non-trivial MPI topology implementation, achieving improvements beyond a factor of five for certain system configurations. Still, the topology functionality lacks expressivity to deal effectively with the mesh-partitioning Problem. A mild extension to MPI is suggested, which, however, still cannot exclude possibly sub-optimal partitioning results. Solving instead the mesh-partitioning Problem outside of MPI requires knowledge of the communication system. We discuss ways in which such could be provided by MPI in a portable way. Finally, we formulate and discuss a more general affinity scheduling Problem.

Orli Waarts - One of the best experts on this subject based on the ideXlab platform.

  • z Amos Fiat x
    2015
    Co-Authors: James Aspnes, Yossi Azar, Serge Plotkin, Orli Waarts
    Abstract:

    In this paper we study the Problem of on-line allocation of routes to virtual circuits (both point-to-point and multicast) where the goal is to route all requests while minimizing the required bandwidth. We concentrate on the case of permanent virtual circuits (i.e., once a circuit is established, it exists forever), and describe an algorithm that achieves an O(logn) competitive ratio with respect to maximum congestion, where n is the number of nodes in the network. Informally, our results show that instead of knowing all of the future requests, it is sucient to increase the bandwidth of the communication links by an O(logn) factor. We also show that this result is tight, i.e. for any on-line algorithm there exists a scenario in which (logn) increase in bandwidth is necessary in directed networks. We view virtual circuit routing as a generalization of an on-line Load Balancing Problem, de ned as follows: jobs arrive on line and each job must be assigned to one of the machines immediately upon arrival. Assigning a job to a machine increases this machine's Load by an amount that depends both on the job and on the machine. The goal is to minimize th

  • On-Line Routing of Virtual Circuits with Applications to Load Balancing and Machine Scheduling
    1997
    Co-Authors: James Aspnes, Yossi Azar, Amos Fiat, Serge Plotkin, Orli Waarts
    Abstract:

    In this paper we study the Problem of on-line allocation of routes to virtual circuits (both point-to-point and multicast) where the goal is to route all requests while minimizing the required bandwidth. We concentrate on the case of permanent virtual circuits (i.e., once a circuit is established, it exists forever), and describe an algorithm that achieves an O(logn) competitive ratio with respect to maximum congestion, where n is the number of nodes in the network. Informally, our results show that instead of knowing all of the future requests, it is sufficient to increase the bandwidth of the communication links by an O(logn) factor. We also show that this result is tight, i.e. for any on-line algorithm there exists a scenario in which \Omega\Gamma137 n) increase in bandwidth is necessary in directed networks. We view virtual circuit routing as a generalization of an on-line Load Balancing Problem, defined as follows: jobs arrive on line and each job must be assigned to one of the ..

  • On-Line Routing of Virtual Circuits with Applications to Load Balancing and Machine Scheduling
    1993
    Co-Authors: James Aspnes, Yossi Azar, Amos Fiat, Serge Plotkin, Orli Waarts
    Abstract:

    In this paper we study the Problem of on-line allocation of routes to virtual circuits (both point-to-point and multicast) where the goal is to route all requests while minimizing the required bandwidth. We concentrate on the case of permanent virtual circuits (i.e., once a circuit is established, it exists forever), and describe an algorithm that achieves an O(logn) competitive ratio with respect to maximum congestion, where n is the number of nodes in the network. Informally, our results show that instead of knowing all of the future requests, it is sucient to increase the bandwidth of the communication links by an O(logn) factor. We also show that this result is tight, i.e. for any on-line algorithm there exists a scenario in which (logn) increase in bandwidth is necessary in directed networks. We view virtual circuit routing as a generalization of an on-line Load Balancing Problem, de ned as follows: jobs arrive on line and each job must be assigned to one of the machines immediately upon arrival. Assigning a job to a machine increases this machine's Load by an amount that depends both on the job and on the machine. The goal is to minimize th