- The paper categorizes hybrid beamforming architectures into full-complexity, reduced-complexity, and virtual-sectorization schemes for massive MIMO systems.
- It details methodologies using both instantaneous and averaged CSI, employing iterative optimization and decoupling strategies to approximate fully digital beamforming performance.
- It explores mm-wave applications by leveraging spatially sparse precoding and selection techniques to balance spectral efficiency with cost and power constraints.
This paper presents a comprehensive survey on hybrid beamforming techniques for massive MIMO systems, authored by Andreas F. Molisch et al. The survey elaborates on various hybrid transceiver architectures, categorizing them based on the required channel state information (CSI), their complexity, and their operational frequency bands. The research focuses on the hybrid combination of analog and digital beamforming to optimize spectral efficiency (SE) while managing cost and power consumption.
Introduction
Massive MIMO technology has significantly contributed to enhancing spectral efficiency by leveraging large antenna arrays at the base station (BS). It allows simultaneous communication with multiple user equipments (UEs). However, the large number of antennas complicates signal processing, increases RF chain costs, and necessitates extensive CSI acquisition. Hybrid transceivers are a promising solution, combining analog preprocessing with digital beamforming to optimize system performance.
The authors categorize hybrid beamforming architectures into three main structures: full-complexity, reduced-complexity, and virtual-sectorization.
- Full-complexity structure (A): Each RF chain connects to all antennas, offering the highest performance but also the highest complexity and cost.
- Reduced-complexity structure (B): Limits each RF chain to connect to a subset of antennas, reducing cost at the expense of some performance.
- Virtual-sectorization structure (C): Partitions the antenna array into sectors, each with its own analog beamformer, minimizing overhead and computational complexity.
When employing instantaneous CSI, the optimal hybrid beamforming matrices are complex to derive due to non-convex optimization problems influenced by the coupling between analog and digital beamformers. Two primary methodologies are:
- Approximating the optimal beamformer: By minimizing the Euclidean distance to the fully digital beamformer, the analog and digital parts are iteratively optimized.
- Decoupling the design of analog and digital beamformers: This simplifies the optimization by assuming methods like MMSE receiver processing, enabling a sequential optimization approach.
Given the significant overhead for obtaining instantaneous CSI, averaged CSI can be utilized for the analog part of the beamforming. Notable approaches include:
- Joint Spatial Division Multiplexing (JSDM): This method groups UEs with similar channel covariance statistics, reducing the training and feedback overhead by using averaged CSI for analog beamforming, followed by digital beamforming using instantaneous CSI.
- Decoupling designs: These focus on simplifying the analog beamformer design by decoupling it from the digital beamformer, optimizing separately based on long-term and instantaneous statistics.
A special hybrid architecture involves a selection matrix that connects the analog beamformer to the best subset of available RF chains, optimizing performance under hardware limitations. Different approaches like antenna and beam selection algorithms enhance performance within practical constraints.
Hybrid beamforming becomes critical at millimeter-wave (mm-wave) frequencies due to high propagation loss and the need for large antenna arrays. Key issues addressed include:
- Exploiting channel sparsity for efficient beamforming, using methods like spatially sparse precoding and beam selection.
- Handling multiple user scenarios with hybrid beamforming, ensuring high SE even under realistic propagation conditions.
- Mitigating the impact of RF hardware imperfections, which become more pronounced at mm-wave frequencies, and exploring the tradeoff between SE and energy efficiency (EE).
Conclusion
The survey concludes by emphasizing that while hybrid beamforming has evolved into a critical technology for massive MIMO systems, its optimal implementation depends on the specific application and channel characteristics. The paper highlights the need for continued research in adapting hybrid beamforming techniques to varying operational conditions, balancing performance and complexity effectively.
The insights provided by this comprehensive survey are invaluable for both theoretical advancements and practical implementations in the field of wireless communications.
