Cellular networks planning: A workload balancing perspective

Cell planning has been a long-standing problem since the very starting commercialization of mobile communications, of which power coverage and capacity coverage are two major objectives. In this paper, we develop a novel cell planning scheme that is effective and efficient for both the conventional cellular systems and the arising heterogeneous networks, e.g., the long term evolution advanced (LTE-A) one. The key idea of our proposal is that we redesign the service regions of base stations (BSs) in a traffic-balanced way, making each BS serve a subregion of ALMOST equal throughput requirement. For a given connected polygonal region to be served by a cellular system, we first divide it into compact and connected subregions based on an infinite optimization formulation while keeping the traffic demands of all subregions as equal as possible. Each subregion will be served by a BS located in it. To avoid yielding ill-shaped subregion that is difficult to be covered by a practical BS, a penalty term is introduced to the objective function and it is required that areas of subregions should not differ too much from each other. Then we select the BS from all candidate sites in each subregion to minimize the total power consumption. By using the proposed dividing and selecting algorithms, we update the boundary of each subregion and the location of each BS in an iterative manner until convergence. Numerical results show that our proposal performs quite well for both randomly generated scenarios and real city environment. The proposed cell planning method provides quality of service (QoS) guaranteed performance with lower capital expenditure and operating expenditure.