Traditional genetic algorithm and random-weighted genetic algorithm with GIS to plan radio network

Cell planning as a process in GSM network planning must take into consideration all the criteria that are related to the cell-planning process, including the technical, financial, and environmental criteria. This paper aims to improve an integrated framework of cell planning that selects the optimal locations of base stations based on finding the trade-offs between coverage area, cost, and the health effects on the populations. The problem of base stations siting in this paper is reformulated from an unconstrained multiobjective optimization problem into a constrained multiobjective optimization problem. Two types of evolutionary algorithms, a traditional genetic algorithm with penalty functions as constraint handling and a random-weighted genetic algorithm (RWGA), were implemented and compared based on which one can produce an optimal solution in less iteration. The minimum number of base stations that are required to cover the conducted area was predicted based on the produced solution and was compared with the number of base stations that are constructed by the Mobinil operator. Cairo was selected as the conducted research area; both models were superior to the Mobinil scenario in covering the conducted area with a less number of base stations.