Accelerating the convergence of a modified Tabu Search algorithm using a new objective function for the frequency assignment problem

The frequency assignment problem (FAP), where the objective is to find the best possible combination that minimizes the total number of violations in an assignment, is studied in this paper using a Tabu Search (TS) algorithm with a dynamic tabu list in order to improve the performance and the effectiveness of original TS algorithm. The basic idea is to really explore and exploit the search space and to escape from local minimum in order to have more chance to find the global optimum. Due to the NP-hardness of the FAP, any optimal solution is guaranteed in a limited time. So that, to reduce the computation time necessary to the convergence of our algorithm, we define a new objective (fitness function) for our model of solution. Our experimental results show that the computation time and accuracy is clearly improved and the proposed TS algorithm can be efficiently applied to find a near-optimal solution.