DNA Algorithm for Solving the Traveling Salesman Problem

Lin He,NA Computer
2006-01-01
Computer Engineering and Applications Journal
Abstract:The traveling salesman problem is given by a set of some cities,and asks for a shortest tour of visiting all cities only once from one city and back to this city.It is a classical NP-complete problem in graph theory.It needs exponential time to solve the problem by the electronic computer.In this paper,the DNA algorithm for the traveling salesman problem is provided by the technology of the molecule biology and the Adleman-Lipton model.It is showed that the problem can be solved in the polynomial time.We use a graph with vertexes and edges to denote these cities and roads.The vertex,weight and direction of the graph are encoded by the DNA strings.Putting these DNA strings into the tube to produce,the solution of the traveling salesman problem can be obtained by a series of basic biology operations.The algorithm is highly parallel.For a traveling salesman problem with n cities,the complicacy of computation is O(n).The innovation of the algorithm is the choice of the city string and weight string’s length,which can reduce the solution of the problem involving in an fixed interval and simplify the computation.
What problem does this paper attempt to address?