Beamforming Optimization for Intelligent Reflecting Surfaces without CSI

This letter considers the communication between a multiple antenna Base Station (BS) and a single antenna user, assisted by an Intelligent Reflecting Surface (IRS). Due to the large number of elements at the IRS, acquiring Channel State Information (CSI) requires many radio-frequency chains and considerable training overhead, what is limiting in practice. We aim to optimize the beamforming at the BS and IRS without CSI, by minimizing the transmit power, subject to a minimum signal-to-noise ratio (SNR). To solve this problem, a new method based on the Particle Swarm Optimization (PSO) technique is proposed. Results show that the new solution achieves similar performance as that of the optimal beamforming with CSI.