Decentralized Beamforming for Multicell System Based on Internal Penalty Function

In this paper, an iterative algorithm for decentralized beamforming with internal penalty function is proposed. This algorithm utilizes the property that a stationary point corresponding to local maximum of the objective function is found when gradient is zero. The convergence of the proposed algorithm is proved. The simulation results show that the proposed algorithm is more robust against intra-cell and inter-cell interference than the traditional ones. Especially the proposed algorithm has higher rates as the number of users increases.