Advertisement

Task Allocation within Mesh Networks: Influence of Architecture and Algorithms

  • Aleksandra PostawkaEmail author
  • Iwona Poźniak Koszałka
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 366)

Abstract

This paper concerns the problem of task allocation within the mesh network. Three mesh structures of processors: rectangle, cylinder and torus, are considered. Two created task allocation algorithms: Generalized Stack-Based Allocation and Generalized Recognition Complete First Fit, are presented and evaluated. The algorithms and structures were tested on different sets of tasks to allocate. The investigations were made using the designed experimentation system and concentrated on the indices of performance such as allocation time, processing time, external fragmentation and latency. Moreover, the influence of the sorting of tasks on the efficiency of algorithms was analyzed. The obtained results allow for some recommendations for the designers of the multiprocessor systems.

Keywords

task allocation mesh network algorithm cylinder torus 

Notes

Acknowledgement

This work was supported by the statutory funds of the Faculty of Electronics, Wroclaw University of Technology, Wroclaw, Poland.

References

  1. 1.
    Zydek, D., Selvaraj, H.: Fast and efficient processor allocation algorithm for torus-based chip multiprocessors. Computers and Electrical Engineering 37, 91–105 (2011)CrossRefzbMATHGoogle Scholar
  2. 2.
    Majkowska, A., Zydek, D., Koszałka, L.: Task Allocation in Distributed Mesh-Connected Machine Learning System: Simplified Busy List Algorithm with Q-Learning Based Queuing. Advances in Intelligent Systems and Computing 226, Springer, 763–772 (2013)Google Scholar
  3. 3.
    Jakimczuk, Ł., Kmiecik, W., Poźniak-Koszałka, I.: Task Allocation Algorithms for 2D Torus Architecture. The Twelfth IARIA International Conference on Networks, 169–174 (2013)Google Scholar
  4. 4.
    Zhu, Y.: Efficient Processor Allocation Strategies for Mesh- Connected Parallel Computers. Journal of Parallel and Distributed Computing 16, 328337 (1992)CrossRefGoogle Scholar
  5. 5.
    Yoo, B.S., Das, Ch.R.: A fast and efficient processor allocation scheme for mesh-connected multicomputers. IEEE Transactions on Computers 51, 46–60 (2002)CrossRefGoogle Scholar
  6. 6.
    Koszałka, L., Kubiak, M., Poźniak-Koszałka, I.: Comparison of SBA - Family Task Allocation Algorithms for Mesh Structured Networks. LNCS, Springer 4331, 21–30 (2006)zbMATHGoogle Scholar
  7. 7.
    Poźniak-Koszałka, I., Proma, W., Koszałka, L., Pol, M., Kasprzak, A.: Task Allocation in Mesh Structure: 2Side LeapFrog Algorithm and Q-learning Based Algorithm, LNCS, Springer 7336, 576–587 (2012)Google Scholar
  8. 8.
    Kaminski, R., Koszałka, L., Poźniak-Koszałka, I., Kasprzak, A.: Evaluation and Comparison of Task Allocation Algorithms for Mesh Networks. Proc. of 9th ICN, IEEE CPS, 104–108 (2010)Google Scholar
  9. 9.
    Geunmo, K., Hyunsoo, Y.: On Submesh Allocation for Mesh Multicomputers: A Best-Fit Allocation and a Virtual Submesh Allocation for Faulty Meshes. IEEE Transactions on Parallel and Distributed Systems 9, 1127–185 (1998)Google Scholar
  10. 10.
    Borowiec G., Postawka A., Koszaka L.: Static task allocation algorithms and influence of architecture on Mesh Structured Networks. Computer Systems Engineering PBW, 66–80 (2014)Google Scholar
  11. 11.
    Dally, W.J.: Performance analysis of k-ary n-cube interconnection networks. IEEE Transactions on Computers 39, 7127–785 (1990)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  • Aleksandra Postawka
    • 1
    Email author
  • Iwona Poźniak Koszałka
    • 2
  1. 1.Faculty of Electronics Computer Architecture GroupWroclaw University of TechnologyWroclawPoland
  2. 2.Faculty of Electronics, Department of Systems and Computer NetworksWroclaw University of TechnologyWroclawPoland

Personalised recommendations