A GPU Accelerated Parallel Heuristic for Travelling Salesman Problem

Travelling Salesman Problem (TSP) is one of the typical NP-hard problems in combinatorial optimization, which is easy to be described but hard to be solved. Its possible amounts of path increase exponentially with the amounts of city. Local search metaheuristics can be used to obtain satisfactory resolution (approximate optimum) in a reasonable time for TSP. However, it is still very CPU time-consuming when solving such a large problem instance. As graphic process units (GPUs) have been evolved to support general purpose computing, they are taken as a major accelerator in scientific and industrial computing. In this paper, we present an optimized parallel heuristic efficiently accelerated on GPUs and the proposed optimization algorithm can provide higher quality solutions within a reasonable computational time for TSP.