A new heuristic method for ring topology optimization: A proposal

Abstract

Nowadays, the ring topology has become a standard in network topology design, at least at the core network. The problem in designing ring topology is how to design the connections of each node to each other with the overall cost of the ring is minimum. This problem is known as the Traveling Salesman Problem (TSP). There are two variables which need consideration in TSP, finding the nodes configuration that has minimum cost, and how long it takes to get this minimum configuration. Brute Force has the capability to find the minimum nodes configuration cost, but with every node added to the overall configuration, the time to solve it is increasing exponentially. While not giving the minimum cost, The Ant Colony Optimization (ACO) algorithm has the advantage of minimal time needed to find the sub-minimum nodes configuration. This paper proposed a new heuristic algorithm to find the sub-minimum nodes configuration. The proposed algorithm has a shorter time, in order of less than 1 second for up to 50 nodes, compared to the Ant Colony Algorithm. Although, the cost of the nodes configuration didn't always lower than the Ant Colony Algorithm.