Optimizing base station location and configuration in UMTS networks

Radio planning and coverage optimization are critical issues when deploying and expanding third generation cellular systems. We investigate mixed integer programming models for locating and configuring base stations in UMTS networks so as to maximize coverage and minimize installation costs. The overall model considers both uplink and downlink directions, that we studied separately in Amaldi et al. (2002, 2003b). The two-stage Tabu Search algorithm we propose exploits solutions of a simplified model for the uplink direction to drastically reduce the computational time required to find good approximate solutions of the overall uplink and downlink model. Computational results obtained for realistic instances are reported and discussed.