Parallel Skeletons for Tabu Search Method Based on Search Strategies and Neighborhood Partition

  • Authors:

  • Maria J. Blesa;Lluís Hernàndez;Fatos Xhafa

  • Affiliations:

  • -;-;-

  • Venue:
  • PPAM '01 Proceedings of the th International Conference on Parallel Processing and Applied Mathematics-Revised Papers
  • Year:
  • 2001

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper we present two parallel skeletons for Tabu Search method -a meta-heuristic for solving combinatorial optimization problems. Our parallel skeletons are designed and implemented from the generic parallel programming paradigm. The first skeleton is based on independent runs model endowed with search strategies; the second one is a master-slave model that uses neighborhood partition. In order to obtain these skeletons, we designed and implemented a sequential skeleton for the method that is used as a basis for the two parallel skeletons. Both skeletons provide the followings: (a) permit to obtain parallel implementations of Tabu Search for concrete problems from existing sequential implementations; (b) there is no need for the user to know neither parallel programming nor communication libraries; (c) the parallel implementations for a concrete problem are obtained automatically from the existing sequential implementation for the problem. The skeletons are implemented in C++ using MPI as a communication library and offer several properties such as a genericity, flexibility, component reuse, and time savings, mainly due to the generic and object oriented programming paradigms. We have instantiated the two skeletons for the 0-1 Multidimensional Knapsack problem and report extensive experimental results.