Planning and optimization of Cellular Networks Using Load-based Voronoi Algorithm

We study the optimal location of Access Points (AP) in cellular networks with irregular service demand distribution, using a power Voronoi algorithm. To this end, we propose two weights updating approaches to acquire uniform load pattern with uniform power allocation at all APs. In Channel-based Power Voronoi Algorithm (CPVA), we update the Voronoi weights based on the approximated loads imposed to APs assuming that the cell load is only a function of traffic and the channel coefficients between the AP and the associated users. In Load-based Power Voronoi Algorithm(LPVA), the weights are updated by load calculation through a convex optimization problem in cell load coupling model. It is worth noting that in both methods, approximately uniform load patterns are derived with uniform power allocation at APs, which guarantees the quality of service and energy efficiency in the network. We also remark that LPVA outperforms CPVA at the cost of higher computational cost. Finally, the results of both method are compared to prior art to show the advantages of applying CPVA and LPVA in cellular network planning.