Full-stack Hybrid Beamforming in mmWave 5G Networks

Felipe Gómez-Cuba, Tommaso Zugno, Junseok Kim, Michele Polese, Saewoong Bahk, Michele Zorzi

While the theoretical potential of millimeter-wave (mmWave) frequencies for 5G is massive, the practical implementation is a complex balancing act of physics, hardware constraints, and protocol logic. This paper by Gómez-Cuba et al. moves beyond idealized "link-level" simulations to explore how Hybrid Beamforming (HBF) behaves when integrated into the full 5G protocol stack, including MAC scheduling and TCP/IP traffic.

The Hardware Dilemma: Analog vs. Digital

MmWave communications offer huge bandwidth but suffer from high path loss. To overcome this, 5G systems use Beamforming (BF) to concentrate signal energy toward specific users. However, hardware designers face a dilemma:

  1. Fully Digital BF provides maximum flexibility but consumes excessive power and is prohibitively expensive for the massive antenna arrays required at mmWave frequencies.
  2. Analog BF is energy-efficient but can only point one beam at a time, effectively limiting the system to Time Division Multiple Access (TDMA), where only one user is served per time slot.

Hybrid Beamforming (HBF) serves as the middle ground, using a limited number of digital chains to control a large number of analog antenna elements. While HBF is well-studied in theoretical literature, this research identifies a critical gap: most studies ignore how HBF interacts with the 3GPP NR protocol stack.

The Pilot Synchronization Trap

A primary contribution of this work is identifying the conflict between "front-loaded" channel estimation and Multi-User MIMO (MU-MIMO) scheduling. In 5G NR, Demodulation Reference Signals (DMRS), or pilots, are sent at the start of a transmission. If User A is already transmitting and User B starts mid-way, User A’s receiver cannot "re-learn" the channel to cancel the new interference introduced by User B.

To solve this, the authors propose the Padded mmWave RR Scheduler (PMRS). This scheduler forces simultaneous start times for users within a "Spatial Division Multiple Access (SDMA) bundle." Even if it requires adding a small amount of "padding" (intentionally idle symbols) to align the transmissions, the resulting interference reduction significantly outweighs the minor loss in resource efficiency.

The Padded mmWave RR Scheduler (PMRS) synchronizes the start times of multiple users to ensure pilot alignment and interference cancellation.

Frequency-Selective MMSE Beamforming (SMBF)

On the physical layer, the authors introduce Frequency-Selective MMSE Beamforming (SMBF). Instead of relying solely on a predefined codebook of analog beams, SMBF adds a low-complexity digital precoding layer. This layer uses a Minimum Mean Squared Error (MMSE) approach to mitigate inter-beam interference across different subcarriers. The researchers argue that joint multi-user beamforming design is imperative; a MU-MIMO system cannot be fully exploited if it is treated as a mere collection of independent single-user analog beams.

Performance and Results

The study utilized a high-fidelity simulation environment based on an extended version of the ns-3 mmWave module. Testing under realistic 3GPP "Urban Macro" scenarios at 28 GHz with TCP/IP traffic, the SMBF+PMRS approach demonstrated significant gains over standard single-layer TDMA and basic Codebook Beamforming (CBF).

Cumulative Distribution Functions (CDF) of SINR and BLER for uplink and downlink, comparing the proposed SMBF+PMRS approach against standard baselines.

Key findings include:

  • Interference Suppression: The SMBF scheme successfully pushed the Signal-to-Interference-plus-Noise Ratio (SINR) of a 4-layer multi-user system to nearly match the performance of a single-user system.
  • Capacity Doubling: In loaded cell scenarios, the proposed full-stack HBF approach achieved a throughput of approximately 450 Mbps in both uplink and downlink. This is roughly double the capacity of standard single-layer approaches (~180-330 Mbps).
  • Reliability: The Block Error Rate (BLER) was significantly stabilized, preventing the outages that typically occur when simple analog beams overlap in dense environments.

Conclusion

The research highlights that for 5G mmWave to reach its full potential, cross-layer design is essential. By synchronizing MAC-layer scheduling with digital physical-layer precoding, network designers can overcome the inherent limitations of analog hardware, effectively doubling system capacity through robust SDMA.

Summary updated

Abstract

This paper analyzes Hybrid Beamforming (HBF) and Multi-User Multiple-Input Multiple-Output (MU-MIMO) in millimeter wave (mmWave) 5th generation (5G) cellular networks considering the full protocol stack with TCP/IP traffic and MAC scheduling. Prior work on HBF and MU-MIMO has assumed full-buffer transmissions and studied link-level performance. We report non-trivial interactions between the HBF technique, the front-loaded channel estimation pilot scheme in NR, and the constraints of MU-MIMO scheduling. We also report that joint multi-user beamforming design is imperative, in the sense that the MU-MIMO system cannot be fully exploited when implemented as a mere collection of single-user analog beams working in parallel. By addressing these issues, throughput can be dramatically increased in mmWave 5G networks by means of Spatial Division Multiple Access (SDMA).
BibTeX
@inproceedings{gomezcuba2021full,
  author = {Felipe Gómez-Cuba and Tommaso Zugno and Junseok Kim and Michele Polese and Saewoong Bahk and Michele Zorzi},
  title = {{Full-stack Hybrid Beamforming in mmWave 5G Networks}},
  booktitle = {MedComNet},
  year = {2021},
  url = {https://openreview.net/forum?id=dfWO1YXLIG},
}
Figure