Annals of Operations Research, 21 (1989) 59-84 59
SERIAL AND PARALLEL SIMULATED ANNEALING AND TABU SEARCH
ALGORITHMS FOR THE TRAVELING SALESMAN PROBLEM
Miroslaw MALEK, Mohan GURUSWAMY, Mihix PANDYA
Department of Electrical and Computer Engineerin~ The University of Texas at Austin, Austin,
Texas 78712, U.S.A.
and
Howard OWENS
Microcomputer Division, Motorola lnc.
Abstract
This paper describes serial and parallel implementations of two different search techniques
applied to the traveling salesman problem. A novel approach has been taken to parallelize
simulated annealing and the results are compared with the traditional annealing algorithm.
This approach uses abbreviated cooling schedule and achieves a superlinear speedup. Also a
new search technique, called tabu search, has been adapted to execute in a parallel computing
environment. Comparison between simulated annealing and tabu search indicate that tabu
search consistently outperforms simulated annealing with respect to computation time while
giving comparable solutions. F, xamples include 25, 33, 42, 50, 57, 75 and 100 city problems.
1. Introduction
The Traveling Salesman Problem (TSP) is a classic problem in combinatorial
optimization research. This is because it is easily defined, and improved al-
gorithms to find optimized solutions to this problem can be adapted to an entire
class of NP-complete problems.
Our objective is to implement combinatorial optimization algorithms such that
they may execute in parallel and exchange data periodically. The goal is to study
the potential performance improvements as compared to a single processor
implementation of the traveling salesman problem.
In particular, we are interested in comparing the time efficiency and cost of
simulated annealing [1] and tabu search [2] algorithms. Each of these algorithms
have the ability to avoid or escape local minima in searching for the global
optimal solution. However, the method of reaching the optimal solution is very
different as will be described later. These tmiversal search algorithms are tested
on a specific heuristic used for the TSP problem.
The traveling salesman problem consists of finding the shortest Hamiltonian
circuit in a complete graph where the nodes represent cities. The weights on the
© J.C. Baltzer A.G. Scientific Publishing Company