基于插值矢量量化的MIMO-OFDMA下行链路预编码

Precoding for MIMO-OFDMA downlinks based on interpolation and vector quantization

为了解决多载波系统预编码需要的反馈数据随载波数线性增长的问题,提出了MI-MO-OFDMA下行链路有限反馈预编码方案。首先,由二元联合矢量量化算法离线构建预编码矩阵的码本,然后,根据载波相关性将载波分成若干频带,由接收端根据容量最大化准则确定每个载波频带的预编码矩阵和所分配的用户等预编码信息,并将信息索引利用有限个比特反馈到发射端,最后,发射端利用索引在码本中找到相应预编码信息,进行插值恢复,得到所有子载波的预编码信息。仿真结果表明,以同样的反馈数据量,所提出的方案可比现有预编码方案获得更大的下行容量。

The feedback data for precoding in multi-carrier systems increases linearly with an increase in the number of subcarriers. To solve this problem, a precoding scheme with limited feedback for MIMO-OFDMA downlinks was proposed. First, a codebook of precoding matrices was constructed offline using the two-variable joint vector quantization algorithm. Then, the subcarriers were divided into subcarrier bands according to the correlation between subcarriers. And the receiver choosed the precoding matrices, the user, the angel parameter for each subcarrier band aiming at the capacity maximization. The indices of the precoding information were conveyed to the transmitter with limited bits of feedback. Finally, the transmitter obtained the information in its codebooks via the indices and then obtains precoding matrices for all the subcarriers by interpolation. Simulation results show that the proposed scheme achieves a larger capacity than the existing schemes with the same number of feedback bits.