Low complexity multicast beamforming in massive MIMO multi-cell networks

Abstract

In this thesis, we consider downlink multicast beamforming in massive multi-input multi- output (MIMO) multi-cell networks. Both non-cooperative and cooperative beamforming scenarios are considered. For the non-cooperative scenario, aiming at maximizing the minimum signal-to-interference-plus-noise (SINR) among users, we propose two multicast beamforming structures: weighted maximum ratio transmission (MRT) and weighted zero-forcing (ZF). Based on the weighted MRT beamforming structure, we propose a multicast beamforming design that transforms the beamforming problem into an optimization problem of weights and solve it via semi-definite relaxation (SDR) and successive convex approximation (SCA). To further reduce the computational complexity and the communication overhead, a distributed weighted MRT beamforming design based on signal-to-leakage ratio (SLR) is developed. Furthermore, we propose a distributed ZF beamforming design to maximize the minimum SINR among users based on the weighted ZF structure. The asymptotically optimal solution for the weighted ZF method with infinite number of antennas is derived. We also extend our work to the cooperative beamforming scenario and develop the weighted MRT approach for cooperative beamforming. Compared with the traditional method which directly solves beamforming problem via SDR approach, our proposed methods have a low computational complexity for massive MIMO systems. Particularly, the computational complexity of weighted MRT methods does not grow with the number of antennas. Therefore our proposed methods are suitable for multi-cell networks equipped with large scale of antennas. Simulation results show that our proposed multicast beamforming solutions yield comparable or better performance than existing approaches but have significantly lower complexity for practical systems with a large but finite number of antennas.