A decomposition method for pilot power planning in UMTS systems

Pilot power management is an important issue for coverage planning in UMTS systems. We consider the problem of minimizing the pilot power subject to the constraint of full service coverage. For this planning problem, which is NP-hard in complexity, effective methods being able to deal with large-scale networks of heterogeneous cell coverage patterns are highly desirable. We propose an integer linear optimization formulation, and a decomposition method that exploits the problem structure using a Dantzig-Wolfe reformulation. We report numerical results for networks of various sizes. The proposed method efficiently finds near-optimal solutions that yield substantial savings in power consumption when compared to baseline approaches.