Fast TSP algorithm based on topological perception

Shuai Zhang,Yun Wang,Kai Li
DOI: https://doi.org/10.3969/j.issn.1001-0505.2014.03.013
2014-01-01
Abstract:To satisfy the characteristics of wireless sensor networks,a high speed heuristic algorithm for traveling salesman problem(TSP)based on topological perception is put forward.Through ana-lyzing the relationship between the topology of wireless sensor networks and the solution of TSP,the topological distance based on the maximum common subgraph isomorphism is defined to compute the similarity of topologies.Then,the input topology is clustered based on the topological distance and the solution of TSP is solved through mapping.In the proposed algorithm,appropriate trim condi-tions are set to improve the speed of the algorithm.A threshold parameter is added to balance the similarity of topologies in class and the number of generated clusters.The simulation results show that when the numbers of the nodes are 90 and 70 in TSP,the running time of this topological per-ception algorithm are 0.615 and 0.508 s,respectively,which are about 3% to 4% of those of the Lin-Kernighan algorithm and the ant cology algorithm.And the precision of the proposed algorithm is between those of the two algorithms.
What problem does this paper attempt to address?