Stochastic control of the scalable high performance distributed computations

  • Authors:
  • Zdzislaw Onderka

  • Affiliations:
  • Dept. of Geoinformatics and Applied Computer Science, AGH, Univ. of Science and Technology, Krakow, Poland

  • Venue:
  • PPAM'11 Proceedings of the 9th international conference on Parallel Processing and Applied Mathematics - Volume Part II
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

The paper presents model desribing the dynamics of the background load for the computers in the network. For that model two problems of the task allocation in the network are formulated: the problem of the stochastic control based on the background load model and the stochastic control based on the Markov Decision Process Theory. The open-loop and closed-loop control algorithms based on the stochastic forecast and Markov Decision Process Theory respectively are presented.