Non-standard cost terminal assignment problems using tabu search approach

Terminal assignment (TA) is important in increasing the telecommunication networks´ capacity and reducing the cost of it. We propose a tabu search (TS) approach to solve the problem with non-standard cost functions. A greedy decoding approach is used to generate the initial solution and then an effective and unique search approach is proposed to produce the neighborhood, which exchange one of the terminals in each concentrator to improve the quality of solution. Simulation results with the proposed TS approach are compared with those using genetic and greedy algorithms. Computer simulations show that our approach achieves very good results in solving this problem.