Using Improved Parallel Ant Colony Optimization Based on Graphic Processing Unit-Acceleration to Solve Motif Finding Problem

摘要

In this paper, we implements on a CUDA-enabled GPU to efficiently perform ensemble simulations of an individual ant colony optimization since the structure of the algorithm are well suited for parallelization; the proposed method has outstanding performance and efficiency compared to the serial CPU and population-based GPU implementations for large scale problems in TSPLIB. More precisely, the simulations show that the stability and the speed of convergence of the improved ACO algorithm can be enhanced greatly. We also use the improved algorithm based on GPGPU in Motif Finding Problem, the results of the simulated experiments show that proposed algorithm can improve the efficiency of the results.