Multicell cooperative base stations in presence of channel estimation errors

Cooperation between multicell base stations is a promising technique that has the potential to mitigate Inter- Cell Interference (ICI) and increase data rates without sacrificing additional frequency spectrum. Classical beamforming techniques have been used in order to eliminate the existing ICI and enhance the capacity of downlink cooperative Multiple Input Multiple Output (MIMO) transmission systems. However, a perfect knowledge of the Channel State Information (CSI) at the transmitter is needed in order to perform an adequate beamforming such as the Block-Diagonalization preprocessing. In practice, this is not the case due to the presence of channel estimation errors. In this paper, we model this multicell cooperative system under the presence of channel estimation errors that are often neglected but that can significantly influence the obtained results. In addition, we consider a general framework where the power that is transmitted from the base stations to the mobile stations is not constrained to be identical. To ensure zero-ICI, the beamforming is deduced from the entire null space spanned by the signals transmitted by the other users. Then, we evaluate the sum rate expression and we propose efficient alternatives in order to improve it without coupling the beamforming and the power control processes. Simulation results are given to support our claims.