摘要
研究了多小区多用户蜂窝网络中子空间干扰对齐技术.与基于预编码与接收滤波器联合设计的传统方法不同,本文提出的方法只需要预编码参与迭代优化,这样能够避免上下行链路反复交替迭代带来的冗余开销.更重要的是我们引入了格拉斯曼流形,将优化问题在此流形上重组,降低了空间维度,提出该流形上的最陡下降沿算法达到子空间干扰对齐.仿真结果表明本文提出的方法比已有算法具有收敛速度快,系统容量高的优势.
The subspace interference alignment for multi-cell and multi-user cellular networks was focused. Different from most previous algorithms that are based on a joint design of precoder and receive filter, the proposed method achieves interference alignment with precoder design only. This means our algorithm only requires the participation of transmitters, which will alleviate significantly the overhead induced by alternation between the up and down links. More importantly, varying from the traditional constrained optimization method, the precoder design on complex Grassmann manifold with lower dimensions was reformulated and a novel steepest descent algorithm was derived to achieve perfect subspace interference alignment. Simulation results suggest that the proposed algorithm has better convergence performance and higher system capacity compared with previous methods.
基金
Supported by National Natural Science Foundation of China(61171112)
MIIT of China(2010ZX03005-001-02,2012ZX03001037,2010ZX03002-010-02)
