A Java implementation of the branch and bound algorithm: The asymmetric traveling salesman problem

Abstract

This paper offers a description of a Java implementation of the branch-and-bound (BnB) algorithm for the Traveling Salesman Problem with asymmetric cost matrix (ATSP). A generic interface for solving minimization problems with BnB is proposed and the results of computational experiments for ATSP with random cost matrices are given for different problem sizes (50,100, 150, 200, 250, and 300 cities). ©JOT, 2005.