Coordinated Multicell Multicast Beamforming Based on Manifold Optimization

Abstract

Multicast beamforming is a key technology for next-generation wireless\ncellular networks to support high-rate content distribution services. In this\npaper, the coordinated downlink multicast beamforming design in multicell\nnetworks is considered. The goal is to maximize the minimum\nsignal-to-interference-plus-noise ratio of all users under individual base\nstation power constraints. We exploit the fractional form of the objective\nfunction and geometric properties of the con-straints to reformulate the\nproblem as a parametric manifold optimization program. Afterwards we propose a\nlow-complexity Dinkelbach-type algorithm combined with adaptive exponential\nsmoothing and Riemannian conjugate gradient iteration, which is guaranteed to\nconverge. Numerical experiments show that the proposed algorithm outperforms\nthe existing SDP-based method and DC-programming-based method and achieves\nnear-optimal performance.\n