Beamforming design of multiple intelligent reflecting surfaces for MIMO communications in LOS channel

Abstract

A way to solve the explosively increasing demand for wireless data traffic is to utilize large bandwidth of highfrequency bands. However, to exploit high-frequency channels such as millimeter wave and terahertz wave bands, high-path loss must be overcome and a large number of multiple antennas must be used, which greatly increases implementation cost, power consumption, and hardware complexity. In particular, the high-frequency channel cannot utilize the rich scattering of the existing microwave bands, and the channel correlation is high, so it becomes a line-of-sight (LOS) channel environment. That is, as it becomes a rank-deficient channel, the spatial multiplexing gain cannot be utilized, thereby limiting the performance of the system. To overcome these problems, intelligent reflecting surfaces (IRS) has been introduced into wireless mobile communication systems. The IRS is a promising technology that can control the incident beam from the transmitting end to improve the performance of the receiving end. In addition, the IRS is cost-effective as it consist of a large number of passive elements that do not require amplification and regeneration. Through this, the IRS can achieve significant performance improvement by increasing the rank of the channel matrix in the high frequency channel environment. In this paper, the IRS beamforming system that supports multi-antenna systems in the LOS environment, which is a rank-deficient channel, is dealt with. To increase the rank of the channel matrix, we propose optimal reflection coefficient design methods for the system with multiple IRSs deployed and analyze that’s performance. Furthermore, we provide the method of joint designing with the IRS by incorporating analog-digital hybrid beamforming technology that can reduce the number of expensive radio frequency chains. Additionally, we provide insight into optimal deployment that can improve performance when designing IRS-assisted systems in real world.