The proceedings contain 69 papers. The special focus in this conference is on distributed Computing. The topics include: Performance of fair distributed mutual exclusion algorithms;a framework for automatic identifica...
ISBN:
(纸本)9783540240761
The proceedings contain 69 papers. The special focus in this conference is on distributed Computing. The topics include: Performance of fair distributed mutual exclusion algorithms;a framework for automatic identification of the best checkpoint and recovery protocol;distributed computation for swapping a failing edge;flexible cycle synchronized algorithm in parallel and distributedsimulation;rule mining for dynamic databases;a novel P2P based e-learn heuristic-based scheduling to maximize throughput of data-intensive grid applications;failure recovery in grid database systems;on design of cluster and grid computing environment toolkit for bioinformatics applications;study of scheduling strategies in a dynamic data grid environment;virtual molecular computing – emulating DNA molecules;complexity of compositional model checking of computation tree logic on simple structures;a multi-agent framework based on communication and concurrency;statistical analysis of a P2P query graph based on degrees and their time-evolution;t-UNITY – a formal framework for modeling and reasoning about timing constraints in real-time systems;finding pareto-optimal set of distributed vectors with minimum disclosure;a fair medium access protocol using adaptive flow-rate control through cooperative negotiation among contending flows in Ad Hoc;an adaptive transmission power control protocol for mobile ad hoc networks;a macro-mobility scheme for reduction in handover delay and signaling traffic in MIPv6;path stability based adaptation of MANET routing protocols;agent-based distributed intrusion alert system;a soft computing intrusion detection system and effect of data encryption on wireless ad hoc network performance.
This paper presents the design and implementation of a reinforcement learning agent that automatically selects appropriate loop scheduling algorithms for parallel loops embedded in time-stepping scientific application...
详细信息
This paper presents the design and implementation of a reinforcement learning agent that automatically selects appropriate loop scheduling algorithms for parallel loops embedded in time-stepping scientific applications executing on clusters. There may be a number of such loops in an application, and the loops may have different load balancing requirements. Further, loop characteristics may also change as the application progresses. Following a model-free learning approach, the learning agent assigned to a loop selects from a library the best scheduling algorithm for the loop during the lifetime of the application. The utility of the learning agent is demonstrated by its successful integration into the simulation of wave packets - an application arising from quantum mechanics. Results of statistical analysis using pairwise comparison of means on the running time of the simulation with and without the learning agent validate the effectiveness of the agent in improving the parallel performance of the simulation.
The proceedings contain 23 papers from the conference on the 18th workshop on parallel and distributedsimulation, PADS 2004. The topics discussed include: simulation validation using direct execution of wireless ad-h...
详细信息
ISBN:
(纸本)0769521118
The proceedings contain 23 papers from the conference on the 18th workshop on parallel and distributedsimulation, PADS 2004. The topics discussed include: simulation validation using direct execution of wireless ad-hoc routing protocols;detailed OFDM modeling in network simulation of mobile ad-hoc networks;event reconstruction in time warp;just-in-time cloning;towards grid-aware time warp;and the effect of detail on ethernet simulation.
Copyright and Reprint Permissions: Abstracting is permitted with credit to the source. Libraries may photocopy beyond the limits of US copyright law, for private use of patrons, those articles in this volume that carr...
Copyright and Reprint Permissions: Abstracting is permitted with credit to the source. Libraries may photocopy beyond the limits of US copyright law, for private use of patrons, those articles in this volume that carry a code at the bottom of the first page, provided that the per-copy fee indicated in the code is paid through the Copyright Clearance Center. The papers in this book comprise the proceedings of the meeting mentioned on the cover and title page. They reflect the authors' opinions and, in the interests of timely dissemination, are published as presented and without change. Their inclusion in this publication does not necessarily constitute endorsement by the editors or the Institute of Electrical and Electronics Engineers, Inc.
If a model shall be executed in a parallel, distributed instead of a sequential manner typically the entire simulation engine has to be exchanged. To adapt the simulation layer more easily to the requirements of a con...
详细信息
ISBN:
(纸本)0769521118
If a model shall be executed in a parallel, distributed instead of a sequential manner typically the entire simulation engine has to be exchanged. To adapt the simulation layer more easily to the requirements of a concrete model to be run in a specific environment a component based simulation layer has been developed for JAMES. A set of different simulator components demonstrates that a component-based design facilitates the exchange of simulators and their combination.
parallelsimulation and distributedsimulation sometimes appear to be two different worlds. Where parallelsimulation aims at increasing the speed of a single model by distributing it over more processors, distributed...
详细信息
ISBN:
(纸本)0769521118
parallelsimulation and distributedsimulation sometimes appear to be two different worlds. Where parallelsimulation aims at increasing the speed of a single model by distributing it over more processors, distributedsimulation looks at ways to link entire models or federates that run on different computers. In the case of distributedsimulation, the models themselves are hard to distribute, and often they each run on one processor as a monolithic model. This paper advocates to build the more traditional simulation models in such a way, that they can be easily distributed. As simulationists, we can learn from component-based theory from the software engineering field to prepare our models for distribution, and parts of our models for reuse. The results of several projects show that componentizing simulation models can have many advantages. The results also show that it is not easy to create models in a componentized way, and that current methods for simulation model building should be adapted.
Time-parallelsimulation offers the potential of massive parallelization of a simulation application, due to the amount of achievable parallelism not being restricted by the decomposability of the state space of a sim...
详细信息
ISBN:
(纸本)0769521118
Time-parallelsimulation offers the potential of massive parallelization of a simulation application, due to the amount of achievable parallelism not being restricted by the decomposability of the state space of a simulation model. Unfortunately, the potential speedup of a time-parallelsimulation highly depends on the ability to match final and initial states of adjacent time intervals. Depending on the properties of the underlying simulation model, it might be feasible to accept a simulation iteration, even if the states of adjacent time intervals do not match exactly. This leads to the concept of approximative state matching in time-parallelsimulation, which is introduced in this paper Experiments with a prototypical implementation of a simple simulation model show encouraging results in terms of simulation speedup and introduced error.
暂无评论