基于波束域降维的低复杂度大规模MIMO波束成形方法

Low complexity beamforming for massive MIMO systems by beam domain dimension reduction

基于迫零准则(Zero Forcing,ZF)的波束成形和基于最小均方误差(Minimum Mean Square Error,MMSE)的波束成形是应用广泛的两种波束成形预编码方式。但是,在大规模MIMO场景下,发送端和接收端的天线数量较大,在应用ZF和MMSE两种预编码方法时,对矩阵的求逆计算变得尤为复杂。文中提出了一种基于波束域降维的波束成形方法来降低预编码计算的复杂度,通过引入波束域信道矩阵的概念,对维度较大的全维信道矩阵进行等效分解,再对分解后的降维矩阵分别做预编码,降低了波束成形预编码计算的复杂度。最后通过仿真验证了该方案的可行性。

Zero forcing （ZF） and minimum mean square error （MMSE） are two widely used methods for beamforming precoding. However, in massive MIMO systems, when the number of antennas at transmit- ter and receiver is larger, the calculation of the inverse matrix becomes more complex. A beam-domain dimension reduction beamforming scheme is proposed by using beam domain channel matrix to decompose the full-dimensional channel matrix and precoding the low-dimensional matrixes, respectively. The com- putational complex of the beamforming precoding is decreased by the scheme. Finally, the feasibility of the scheme is verified by simulation.