Hybrid Fire Hawk With Successive Convex Approximation for Sum Rate Maximization Problem in Double IRS-Assisted Multi-User MIMO mmWave Systems

Abstract

A promising technique for improving spectral efficiency in wireless communication is Intelligent Reflecting Surfaces (IRS). However, because of significant path loss and obstructions, a single IRS is not enough to provide adequate coverage and beamforming gain in millimeter-wave (mmWave) networks. To overcome these limitations, this paper investigates the impact of a dual-IRS-assisted multi-user MIMO mmWave system, which enables cooperative passive beamforming to enhance the effective channel gain and extend coverage in non-line-of-sight (NLoS) environments. The proposed approach optimizes phase shift design at the IRSs and digital precoding at the transmitter by formulating a weighted sum rate maximization issue. To effectively solve the precoding and phase shift design problem, a hybrid metaheuristic optimization framework that combines Bernstein-Levy Search Differential Evolution (BL-SDE), Hybrid Aquila with Fire Hawk (HAOFH) optimization, and Double Stochastic Successive Convex Approximation (DSSCA) is generated. The hybrid Aquila optimizer specifically solves the digital precoding matrix design challenge, while the Fire Hawk optimizer solves the analog phase shift problem. Throughput maximization is a critical indicator for assessing IRS-assisted mmWave MIMO systems, and its direct impact on network efficiency and user experience is the driving force for its adoption as the performance metric. According to simulation results, the suggested dual-IRS system outperforms traditional single-IRS and non-IRS-assisted schemes in terms of spectral efficiency, sum rate, bit error rate, and mean square error. These findings support the efficiency of the dual-IRS framework in addressing mmWave channel defects and promoting next-generation wireless communication.