Optimizing least-cost steiner tree in graphs via an encoding-free genetic algorithm

Abstract

Most bio-inspired algorithms for solving the Steiner tree problem (STP) require the procedures of encoding and decoding. The frequent operations on both encoding and decoding inevitably result in serious time consumption and extra memory overhead, and then reduced the algorithms’ practicability. If a bio-inspired algorithm is encoding-free, its practicability will be improved. Being motivated by this thinking, this article presents an encoding-free genetic algorithm in solving the STP. To verify our proposed algorithm’s validity and investigate its performance, detailed simulations were carried out. Some insights in this article may also have significance for reference when solving the other problems related to the topological optimization.