Abstract

Cell planning includes planning a network of base stations providing a coverage of the service area with respect to current and future traffic requirements, available capacities, interference, and the desired quality-of-service. This paper studies cell planning under budget constraints through a very close-to-practice model. This problem generalizes several problems such as budgeted maximum coverage, budgeted unique coverage, and the budgeted version of the facility location problem. We present the first study of the budgeted cell planning problem. Our model contains capacities, non-uniform demands, and interference that are modeled by a penalty-based mechanism that may reduce the contribution of a base station to a client as a result of simultaneously covering this client by other base stations. We show that this very general problem is NP-hard to approximate and thus we define a restrictive version of the problem that covers all interesting practical scenarios. We show that although this variant remains NP-hard, it can be approximated within of the optimum. a factor of (e−1)/(2e−1)

Citations