SparkACO: A Spark-Based Ant Colony Optimization for Path Planning of Large-Scale Complex Road Networks
Qibu Xiang,Shuxiang Xu,Yushun Fan,Jia Zhang
DOI: https://doi.org/10.1109/bigdataservice58306.2023.00038
2023-01-01
Abstract:Path Planning is a venerable and ubiquitous problem in graph theory. With the development of Internet technology and satellite navigation systems, Path Planning has become an indispensable service. However, along with technological advances, there has been an explosion in demand and data volume. Existing solutions (e.g. Dijkstra algorithm) for Path Planning typically have high computational complexity, and thus it is nontrivial to achieve fast Path Planning in complex networks. To this end, we propose a distributed Ant Colony Optimization based on the Spark engine (SparkACO) to efficiently solve the Path Planning problems in large-scale complex networks. Specifically, we improve the Ant Colony Optimization, a heuristic optimization algorithm, to take full advantage of its speed-up property on optimization problems and apply it to solving Path Planning problems. In addition to this, based on the hierarchical optimization framework, we design a two-layer network with high-speed paths and common paths and then realize the Path Planning after hierarchical processing. For the environment configuration, we construct a computer cluster based on the Hadoop framework and deploy the Spark computing engine as the carrier of the algorithm. Extensive experimental results on a publicly available road network dataset demonstrate the feasibility and superiority of our proposed method.