Robust beamforming designs for downlink cloud radio access networks

Abstract

This paper addresses the robust beamforming design for a downlink cloud radio access network (cloud RAN) subject to per-base-station (BS) power constraint and individual signal-to-interference-plus-noise ratio (SINR) requirements. Our objective is to minimize the overall network power and backhaul cost while guaranteeing the users' SINR constraints for every channel realization by designing the joint beamforming vectors robust to the channel uncertainty. Two fundamental downlink transmission strategies named data-sharing strategy and compression strategy are taken into account. In order to solve the non-convex optimization problem, we model the total power by l 0 /l 2 -norm functions and use the semidefinite relaxation (SDR) and l 0 -norm approximation to transform the original problem into tractable ones. Then, we utilize the worst-case approach to deal with the CSI uncertainty and use the majorization-minimization (MM) algorithm to solve the transformed optimization problem. Simulation results verify that the proposed robust designs can significantly increase the output SINR and deal with the channel uncertainty.