Experimental Study of Channel Estimation and Transmit Diversity for IRS-based MISO system

An intelligent reflecting surface (IRS) is an alternative antenna array architecture capable of different applications. By adjusting the phase shift at each IRS unit cell, the incidence signal can be reflected with desired properties to establish qualified wireless communication links. Therefore, it is essential to derive the channel state information (CSI) via each IRS unit cell as accurately as possible. In this paper, the feasibility of channel estimation for an IRS-based multiple-input multiple-output hybrid beamforming system is experimentally studied. Interleaved orthogonal frequency division multiplexing scheme is implemented to efficiently derive the CSI for different IRS sub-arrays. Then, the strongest channel impulse response maximization is further applied to reduce the computational effort for the IRS unit cell phase shift determination. Measurements results are gathered on an IRS testbed operating at 28GHz. In comparison to the conventional beam training algorithm, the proposed channel estimation procedure provides better accuracy, which leads to higher signal-to-noise ratios with the help of transmit diversity combining technique.