Is Active IRS Useful for mmWave Wireless Networks or Not?

Abstract

Recently, intelligent reflecting surfaces (IRSs) have been proposed to enhance the system performance of several applications via smartly reconfiguring signal propagation. Precisely, as low-cost passive devices, IRSs control the scattering, refraction, and reflection characteristics of the radio waves to suppress the interference at one or more desired receivers. In this paper, it is assumed that an active IRS-assisted multiple-input single-output (MISO) system aids a millimeter-wave (mmWave) wireless network. So, the signal sent from the access point (AP) could be reflected by the IRS and simultaneously be received by the user. To achieve the full gain of IRS-assisted wireless networks, not only the phase shift at the IRS should be optimized, but the amplitude variation with low-cost hardware at the IRS should also be exploited. Thus, in this paper, we formulate the system sum rate maximization problem to optimize both the active and passive beamformer at the AP and the IRS, considering amplitude control at the IRS from the limited available power with low-cost hardware. To do so, two low-complexity algorithms are developed. In particular, the first subproblem is solved based on the weighted minimum square error (WMMSE) to optimize the beamforming at the AP. The second sub-problem is solved via successive convex approximation. Numerical results show that active IRS is beneficial and achieves better gains as compared to passive IRS.