摘要
A joint beamforming algorithm is proposed for intelligent reflecting surface(IRS) aided wireless multiple-input multiple-output(MIMO) communication using statistical channel state information(CSI). The beamforming is done by alternatively optimizing the IRS reflecting coefficients and the covariance matrix of the transmit symbol vector, such that the ergodic rate of the system is maximized. The algorithm utilizes only the second order momentum of the random channel matrices and does assume any specific channel distribution, leading to a general framework for ergodic rate evaluation. A practical channel correlation model is configured to validate the performance gain. It is found that the rate can be enlarged by the joint optimization algorithm, however, the gain over that of randomly deployed reflecting coefficients depends highly on the relative correlation distance of the IRS elements and the spatial position of the IRS. In particular, the results suggest that IRS should be placed in the vicinity of either the transmitter or the receiver. Placing IRS far away from those positions is non-beneficial.
基金
supported by the National Key R&D Program of China under grant 2018YFB1801101 and 2016YFB0502202
Zhejiang Lab(No.2019LC0AB02),NSFC projects(61971136,61601119,61960206005,and 61803211)
Jiangsu NSF project(No.BK20191261)
the Fundamental Research Funds for the Central Universities
Young Elite Scientist Sponsorship Program by CAST(YESS20160042)
Zhishan Youth Scholar Program of SEU。
