一种基于自适应子空间跟踪与Jacobi算法的时变信道有限反馈预编码方法

A limited feedback precoding method based on adaptive subspace tracking and Jacobi algorithm for time-varying channels

提出了一种基于自适应子空间跟踪与Jacobi算法的时变信道有限反馈预编码方法。该方法针对无线信道的时变性,利用梯度算法自适应跟踪时变信道的发射预编码矩阵,同时考虑了反馈信道有限比率的反馈比特数,将Jacobi迭代算法应用于时变信道右奇异矩阵的跟踪,从而有效降低了每个反馈帧的反馈比特数。计算机仿真表明,在时变信道,该方法在降低反馈量的同时,还获得了比存在反馈延时的Grassmannian预编码方法更好的系统容量性能。

This paper proposes a limited feedback preceding method based on adaptive subspace tracking and Jacobi algorithm for time-varying channels. In view of the time-variability of wireless channels, this method uses a gradient algorithm for transmission tracking of the precoding matrix for time-varying channels. Meanwhile, considering the bandwidth limits of feedback channels, it uses the Jacobi algorithm to update the right singular matrix of time-varying channels, which significantly reduces the feedback bits. The simulations show that the new adaptive precoding method not only reduces the feedback bits, but also results in larger system sum capacity gains than the Grassmannian precoding method with feedback delay.