Comparing Parallelization of an ACO: Message Passing vs. Shared Memory

Abstract

Abstract. We present a shared memory approach to the parallelization of the Ant Colony Optimization (ACO) metaheuristic and a performance comparison with an existing message passing implementation. Our aim is to show that the shared memory approach is a competitive strategy for the parallelization of ACO algorithms. The sequential ACO algorithm on which are based both parallelization schemes is first described, followed by the parallelization strategies themselves. Through experiments, we compare speedup and efficiency measures on four TSP problems varying from 318 to 657 cities. We then discuss factors that explain the difference in performance of the two approaches. Further experiments are presented to show the performance of the shared memory implementation when varying numbers of ants are distributed among the available processors. In this last set of experiments, the solution quality obtained is taken into account when analyzing speedup and efficiency measures.

Parallel Ant Colony Optimization: Algorithmic Models and Hardware Implementations

“…Delisle et al [5,6] implemented this scheme on shared-memory architectures like SMP computers and multi-core processors. They also compared performance between the two types of architectures [7].…”

“…Emerging from several years of work by the authors on the parallelization of ACO in various computing environments including clusters, symmetric multiprocessors (SMP), multicore processors and graphics processing units (GPU) [5][6][7][8][9][10], it is based on the concepts of computing entities and memory structures. It provides a conceptual vision of parallel ACO that we believe more balanced between theory and practice.…”

Fine-grained Parallel Ant Colony System for Shared-Memory Architectures

“…Finegrained methods focus on completing a single ACS process as fast as possible, by taking the advantage of parallel computing solutions [7]. This is usually done by running a subset of ants in each processing unit, and exchange some sort of information between them [8]. On the other hand, the coarsegrained methods, also known as Multi-instance or Colonybased methods, try to run multiple instances of ant colony systems concurrently [9][10][11].…”

“…Recent parallel ACS proposals for shared memory systems have chosen the coarse-grained model. Delisle et al [8] have compared these methods with respect to varying parameters, such as the number of ants, the problem size, etc.…”

Very Strongly Constrained Problems: An Ant Colony Optimization Approach

“…As reported in Table 1, current works on ACO are devoted to exploit the implicit parallelism of the algorithm in order to speed up the computation on modern multi-core processors. While these studies are focused on particular problems, we can recognize common strategies aimed at splitting the problem (Ouyang and Yan, 2004), exchanging information among colonies (Ellabib et al, 2007), and deploying the best implementation (Delisle et al, 2005). In the following, we will outline some of the most relevant applications.…”