A Type-2 Fuzzy Based Multi-Objective Optimisation for Strategic Network Planning in the Telecommunication Domain

Strategic planning within the telecommunication domain has become necessary to ensure that networks supplying broadband are capable of being designed and updated proactively in order meet the requirements of societies which are increasingly reliant on it. However, these networks are mired in many, often contradictory, objectives. As such these networks and other similar real-world problems fall within the domain of multiobjective optimisation. Multi-objective optimisation performs well within strategic planning where no single solution is sought; instead, a set of non-dominated solutions belonging to a pareto front can be explored. The formation of the pareto front is dependent on the evaluation of solutions dominance over one another. However, traditional dominance evaluation struggles to produce a pareto front that has not become saturated with nonoptimal solutions as complexity increases, due to the number of objectives and the uncertainty that surrounds real-world problems. In this paper, we will present an interval type-2 fuzzy logic system for determining the dominance between solutions in a multi-objective problem. We have validated the performance of the proposed interval type-2 fuzzy logic system in the optimisation of a standard test problem and a real-world business problem. We have transposed the type-2 fuzzy logic system in the place of the standard dominance evaluation system of the NSGA-II algorithm and found that the proposed system outperforms the traditional and type-1 fuzzy dominance rules in both convergence and diversity of solutions towards the true pareto front.