Abstract:Hybrid beamforming is an emerging technology for massive multiple-input multiple-output (MIMO) systems due to the advantages of lower complexity, cost, and power consumption. Recently, intelligent reflection surface (IRS) has been proposed as the cost-effective technique for robust millimeter-wave (mmWave) MIMO systems. Thus, it is required to jointly optimize a reflection vector and hybrid beamforming matrices for IRS-aided mmWave MIMO systems. Due to the lack of RF chain in the IRS, it is unavailable to acquire the TX-IRS and IRS-RX channels separately. Instead, there are efficient methods to estimate the so-called effective (or cascaded) channel in literature. We for the first time derive the near-optimal solution of the aforementioned joint optimization only using the effective channel. Based on our theoretical analysis, we develop the practical reflection vector and hybrid beamforming matrices by projecting the asymptotic solution into the modulus constraint. Via simulations, it is demonstrated that the proposed construction can outperform the state-of-the-art (SOTA) method, where the latter even requires the knowledge of the TX-IRS and IRS-RX channels separately. Furthermore, our construction can provide robustness for channel estimation errors, which is inevitable for practical massive MIMO systems.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is how to jointly optimize the reflection vector and the hybrid beamforming matrix in a millimeter - wave intelligent reflecting surface - assisted large - scale MIMO system to improve the spectral efficiency and robustness of the system. Specifically, since the intelligent reflecting surface (IRS) lacks radio - frequency links and cannot obtain the channel information from the transmitter to the IRS (TX - IRS) and from the IRS to the receiver (IRS - RX) separately, but can only estimate the so - called effective channel. Therefore, the paper proposes a method to achieve approximately optimal reflection vector and hybrid beamforming matrix using only the effective channel.
### Main contributions:
1. **Relaxing the modulus constraint**: By relaxing the modulus constraint, a relaxed optimization problem is defined, and its solution can provide the upper bound of the maximum spectral efficiency. Using this relaxation, the optimal solutions of the relaxed problem can be obtained, including the relaxed reflection vector, analog precoder and combiner, and digital precoder and combiner. It is theoretically proven that in the large - scale system limit, these solutions satisfy the modulus constraint and become approximately optimal solutions to the original problem.
2. **Approximately optimal reflection vector**: By performing singular value decomposition (SVD) on the effective channel, an approximately optimal reflection vector is obtained. It is theoretically proven that at high signal - to - noise ratio (SNR), the gap between this reflection vector and the optimal performance is negligible.
3. **Practical construction method**: Based on theoretical analysis, the practical reflection vector and hybrid beamforming matrix are constructed by projecting the approximately optimal solutions onto the modulus constraint. Simulation results show that the proposed method outperforms existing methods in large - scale but finite millimeter - wave SU - MIMO systems and is robust to channel estimation errors.
### System model:
The paper considers a narrow - band millimeter - wave MIMO system, in which the IRS is equipped with multiple reflecting elements to assist the communication between the transmitter (TX) and the receiver (RX). TX and RX use a limited number of radio - frequency links respectively to implement hybrid beamforming to reduce complexity, cost, and power consumption. The total channel response consists of the TX - IRS channel, the IRS - RX channel, and the direct TX - RX channel.
### Key techniques:
- **Effective channel estimation**: Use a low - complexity method to estimate the effective channel instead of estimating the TX - IRS and IRS - RX channels separately.
- **Relaxed optimization problem**: By relaxing the modulus constraint, the non - convex optimization problem is transformed into a solvable optimization problem.
- **Theoretical analysis**: Through theoretical analysis, it is proven that the proposed method is approximately optimal in the large - scale system limit.
- **Practical application**: Through the projection operation, the theoretical solutions are transformed into practically usable reflection vectors and beamforming matrices.
### Simulation results:
Simulation results show that the proposed method outperforms existing methods in large - scale but finite millimeter - wave SU - MIMO systems and is robust to channel estimation errors. In addition, the proposed method almost achieves the ideal performance with lower training overhead.
In conclusion, through a series of theoretical analyses and practical construction methods, this paper solves the problem of jointly optimizing the reflection vector and the hybrid beamforming matrix in a millimeter - wave intelligent reflecting surface - assisted large - scale MIMO system, providing an effective solution for improving the spectral efficiency and robustness of the system.