Abstract:The optimization of cooperative beamforming vectors in cell-free massive MIMO (mMIMO) systems is presented where multi-antenna access points (APs) support downlink data transmission of multiple users. Albeit the successes of the weighted minimum mean squared error (WMMSE) algorithm and their variants, they lack careful investigations about computational complexity that scales with the number of antennas and APs. We propose a generalized and reduced WMMSE (G-R-WMMSE) approach whose complexity is significantly lower than conventional WMMSE. We partition the set of beamforming coefficients into subvectors, with each subvector corresponding to a specific AP. Such a partitioning approach decomposes the original WMMSE problem across individual APs. By leveraging the Lagrange duality analysis, a closed-form solution can be derived for each subproblem, which substantially reduces the computation burden. Additionally, we present a parallel execution of the proposed G-R-WMMSE with adaptive step sizes, aiming at further reducing the time complexity. Numerical results validate that the proposed G-R-WMMSE schemes achieve over 99% complexity savings compared to the conventional WMMSE scheme while maintaining almost the same performance.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to efficiently optimize the cooperative beamforming vectors in cell - free massive MIMO (massive Multiple - Input Multiple - Output) systems to address the challenges in computational complexity of the traditional Weighted Minimum Mean - Square Error (WMMSE) algorithm. Specifically, the paper proposes a Generalized Reduced - WMMSE (G - R - WMMSE) method, aiming to significantly reduce the computational complexity while maintaining almost the same performance as the traditional WMMSE algorithm. ### Problem Background In cell - free massive MIMO systems, multiple multi - antenna Access Points (APs) support the downlink data transmission of multiple users. Although the WMMSE algorithm and its variants perform well in optimizing beamforming vectors, their computational complexity increases significantly with the increase in the number of antennas and APs, which makes them impractical in practical applications. ### Solutions Proposed in the Paper To solve the above - mentioned problems, the paper proposes the G - R - WMMSE method, whose main features include: 1. **Decomposition of Beamforming Coefficients**: Divide the set of beamforming coefficients into sub - vectors, and each sub - vector corresponds to a specific AP. This partitioning method can decompose the original WMMSE problem into multiple sub - problems for a single AP. 2. **Lagrangian Dual Analysis**: Utilize Lagrangian dual analysis to derive the closed - form solutions of each sub - problem, thereby greatly reducing the computational burden. 3. **Parallel Execution Strategy**: Introduce a parallel execution strategy with an adaptive step size to further reduce the time complexity. Through parallel computing, the beamforming vectors can be updated simultaneously on multiple APs. ### Main Contributions - **Significantly Reduced Complexity**: Numerical results verify that the proposed G - R - WMMSE scheme achieves more than 99% complexity savings compared to the traditional WMMSE scheme. - **Unchanged Performance**: While significantly reducing the complexity, the G - R - WMMSE scheme can maintain almost the same performance as the traditional WMMSE scheme. - **High Adaptability**: This method is applicable to cell - free massive MIMO systems with per - AP power constraints, expanding the application range of the traditional R - WMMSE algorithm. ### Numerical Results The paper verifies the effectiveness of the proposed method through simulation. The results show that under different conditions (such as the number of APs, SNR, etc.), the G - R - WMMSE scheme not only significantly reduces the computational complexity but also maintains a performance comparable to that of the traditional WMMSE scheme and is superior to other benchmark methods (such as Zero - Forcing (ZF) and Maximum Ratio Transmission (MRT)). ### Conclusion The G - R - WMMSE method proposed in the paper provides a new solution for the efficient beamforming optimization of cell - free massive MIMO systems, significantly reducing the computational complexity while maintaining excellent performance. This result is of great significance for the optimal design of future wireless communication systems.

Generalized Reduced-WMMSE Approach for Cell-Free Massive MIMO With Per-AP Power Constraints

Latency Minimization for Mmwave D2D Mobile Edge Computing Systems: Joint Task Allocation and Hybrid Beamforming Design.

Energy Efficient Power Allocation for Cell-Free mmWave Massive MIMO With Hybrid Precoder

Cell-free ISAC Massive MIMO Systems with Capacity-Constrained Fronthaul Links

Energy Efficient Beamforming for User-Centric Virtual Cell Networks.

On the Energy Efficiency of Limited-Backhaul Cell-Free Massive MIMO

Joint Beamforming Design and Stream Allocation for Non-Coherent Joint Transmission in Cell-Free MIMO Networks

Joint User Association and Power Control for Cell-Free Massive MIMO

On the Total Energy Efficiency of Cell-Free Massive MIMO

ADMM for Downlink Beamforming in Cell-Free Massive MIMO Systems

Joint Power Allocation and Hybrid Beamforming for Cell-Free mmWave Multiple-Input Multiple-Output with Statistical Channel State Information

Cell-Free Massive MIMO versus Small Cells

Iterative Access Point Selection, MMSE Precoding and Power Allocation for Cell-Free Networks

6DMA-Aided Cell-Free Massive MIMO Communication

WMMSE-Based Joint Transceiver Design for Multi-RIS Assisted Cell-free Networks Using Hybrid CSI

Bandwidth Efficiency Maximization for Single-Cell Massive Spatial Modulation MIMO: an Adaptive Power Allocation Perspective

Joint AP-UE Association and Precoding for SIM-Aided Cell-Free Massive MIMO Systems

Joint Power and User Grouping Optimization in Cell-Free Massive MIMO Systems

Multiuser Beamforming for Partially-Connected Millimeter Wave Massive MIMO

Efficient Angle-Domain Processing for FDD-based Cell-free Massive MIMO Systems