Sum rate maximization for mm-wave multi-user hybrid IRS-assisted MIMO systems

Abstract

The employment of Intelligent Reflecting Surface (IRS) in MIMO systems has been extensively studied as a feasible alternative to relays. IRS can operate either actively or passively. In the case of a passive IRS configuration, no signal processor or amplifier is required, allowing for signal reflection in the desired phase shift with lower power consumption and cost. However, the passive IRS gain is significantly decreased due to the path loss in the IRS channel. In the active IRS structure, each reflective element is fitted with a power amplifier that enables it to reflect the signal to the intended destination at an appropriate power level. This confers a greater benefit to the system than a passive structure. Consequently, a hybrid active–passive IRS is recommended as it results in a higher sum rate, more energy gain, and a balance between the higher rate and lower energy consumption. In this paper, we examine a practical Rician channel hybrid IRS-assisted multiuser MIMO system with the goal of maximizing the sum rate. To accomplish this, we simplified the objective function of the main problem using the Fractional Programming (FP) method and modified the formulations in order to divide the main problem into a series of sub-problems. Then, using the Block Coordinate Descent (BCD) method and placing sub-problems into consecutive blocks, we solve each sub-problem to obtain the optimal solution for each block and, ultimately, the answer to the main problem. Finally, simulations were conducted for the proposed scenario, indicating that the sum rate of the hybrid IRS structure varied between the active and inactive IRS structure, depending on transmission power. However, adjusting the number of elements in the active and passive structures can achieve higher overall rates than traditional structures at low power levels.