A genetic algorithm for the two-level location area planning in telecommunication networks

With the growing number of mobile users, the management of wireless networks to expand network capacity is of increasing importance. The Location Areas Planning (LAP) problem examines the redistribution of network resources to prevent any degradation in the quality of service. The optimization of these resources minimizes the cost of the registration signaling generated by the procedures of mobility. This paper introduces a new formulation of the “Bi-objective Location Area-Planning” (BOLAP) problem. Modeled as a Two-level assignment problem, we minimize two objectives successively: the location Update (LU) then the cost of the BTS-BSC links connexions. The model is then iterated until no more improvement is performed in the set of optimal solutions. We propose to adopt a Genetic Algorithm (GA) to solve the Two-level BOLAP model. Applied on a sample of real instances with different sizes for a big Tunisian telephony operator, our method generates a set of potentially efficient solutions of a good quality in a practicable CPU time. A comparison with the vector evaluated particle swarm optimization (VEPSO) is also reported.