Millimeter-Wave NR-U and WiGig Coexistence: Joint User Grouping, Beam Coordination, and Power Control

Xiaoxia Xu, Qimei Chen, Hao Jiang, Jun HuangIEEE Trans. Wireless Communications 2022 · DOI

TL;DR

This paper proposes a novel framework for the coexistence of 5G New Radio in Unlicensed (NR-U) and Wireless Gigabit (WiGig) networks operating in the 60 GHz millimeter-wave band. It introduces a joint optimization strategy for user grouping, beam coordination, and power control to maximize NR-U spectral efficiency while ensuring WiGig performance and meeting NR-U latency requirements. The proposed solution utilizes a Lyapunov optimization-based approach and a penalty dual decomposition (PDD) method with a concave-convex procedure (CCCP) to tackle the complex optimization problem.

What Problem Does This Paper Address?

The increasing demand for high data rates and connectivity in future wireless networks is driving the adoption of millimeter-wave (mmWave) frequencies, particularly for 5G New Radio in Unlicensed (NR-U) bands. However, these bands are already occupied by incumbent Wireless Gigabit (WiGig) networks. Ensuring seamless coexistence between NR-U and WiGig is crucial to avoid interference and maximize spectrum utilization. Existing solutions for mmWave coexistence often rely on traditional listen-before-talk (LBT) mechanisms, which are not well-suited for the highly directional beams and complex MIMO systems used in NR-U and WiGig. Furthermore, achieving high spectral efficiency for NR-U while guaranteeing low latency and preventing performance degradation for WiGig users in a shared mmWave spectrum is a significant challenge. This paper aims to address this by developing an integrated approach that jointly optimizes multiple aspects of the network operation.

Methods

The paper proposes a comprehensive framework that integrates several key techniques:

  • MIMO-NOMA for NR-U: The core of the NR-U network utilizes Multiple-Input Multiple-Output Non-Orthogonal Multiple Access (MIMO-NOMA) to serve multiple users simultaneously on the same frequency resources, enhancing spectral efficiency.
  • Hybrid Beamforming: Both NR-U and WiGig networks employ hybrid beamforming, which combines analog and digital beamforming to achieve high directivity while managing RF chain complexity.
  • Lyapunov Optimization: To handle the dynamic nature of wireless networks and the trade-off between NR-U latency and WiGig interference, a Lyapunov optimization framework is used. This approach transforms an infinite-horizon optimization problem into a tractable single-frame problem by introducing virtual queues to track queue backlogs and interference levels.
  • Penalty Dual Decomposition (PDD) and Concave-Convex Procedure (CCCP): The resulting optimization problem is a complex mixed-integer nonlinear programming (MINLP) problem with nonconvex and unit-modulus constraints, making it NP-hard. To solve this, the authors propose a PDD-CCCP method. This involves:
  • Transforming the MINLP into an augmented Lagrangian (AL) problem.
  • Using the CCCP to approximate the nonconvex parts of the AL problem.
  • Decomposing the approximated AL problem into a series of nested convex subproblems using an inexact block coordinate update (BCU) method.
  • Iteratively solving these subproblems and updating dual variables and penalty parameters.
  • Queue-Aware Decoding Order: A novel approach is introduced where the successive interference cancellation (SIC) decoding order for NR-U users is determined based on their queue backlogs, prioritizing users with longer queues to improve their data rates.

Datasets

Not discussed in the paper. The evaluation is based on simulations with defined system parameters.

Key Insights and Ideas

  • Coordinated Beamforming is Key: The paper highlights that effective coexistence in mmWave bands requires not just interference management but active coordination of highly directional beams between NR-U and WiGig networks.
  • Joint Optimization is Necessary: Treating user grouping, beam coordination, and power control as independent problems is suboptimal. Jointly optimizing these aspects leads to significant performance gains.
  • Lyapunov Optimization for Dynamic Trade-offs: The use of Lyapunov optimization provides a robust framework to dynamically balance competing objectives like NR-U spectral efficiency, latency, and WiGig interference, adapting to changing network conditions.
  • PDD-CCCP for Tractability: The proposed PDD-CCCP method is a crucial contribution, enabling the solution of a highly complex, non-convex optimization problem that would otherwise be intractable.
  • Queue-Aware SIC for Enhanced Fairness: Prioritizing users with larger queue backlogs for SIC decoding is an innovative way to improve fairness and data rates for users experiencing higher latency.

Key Results

The paper demonstrates the effectiveness of the proposed PDD-CCCP algorithm through simulations:

  • Spectral Efficiency: The proposed scheme achieves the highest NR-U spectral efficiency compared to baseline schemes (CHS-HBF-FP, CHS-HBF-EP, CHS-HBF-AO, CHS-BCU-SCA).
  • Traffic Delay: It also achieves the lowest NR-U traffic delay, especially as the number of NR-U users increases.
  • Inter-RAT Interference: The proposed method effectively mitigates inter-RAT interference to WiGig networks, keeping it at a manageable level.
  • Convergence: The PDD-CCCP algorithm converges within a reasonable number of iterations (less than 100 inner iterations for spectral efficiency).
  • Trade-off Control: The influence of Lyapunov control parameters (V₁) is shown to dynamically control the trade-off between NR-U traffic delay and WiGig interference.

Limitations and Future Work

  • Computational Complexity: While the PDD-CCCP method makes the problem tractable, the computational complexity is still significant, as detailed in the paper.
  • Assumptions: The model relies on certain assumptions, such as perfect channel state information (CSI) estimation for NR-U users and WiGig devices, and the availability of a pre-defined WiGig analog beamforming codebook.
  • Future Work: The authors suggest that the model formulation can be extended to incorporate uplink and downlink WiGig MU-MIMO transmissions. Further research could explore more advanced CSI acquisition techniques and adaptive codebook designs. The paper also mentions that the LBR mechanism has the highest traffic delay, suggesting that further improvements could be made to LBR or alternative coexistence mechanisms.

Summary updated

BibTeX
@article{xu2022millimeter,
  author = {Xiaoxia Xu and Qimei Chen and Hao Jiang and Jun Huang},
  title = {{Millimeter-Wave NR-U and WiGig Coexistence: Joint User Grouping, Beam Coordination, and Power Control}},
  journal = {IEEE Trans. Wireless Communications},
  year = {2022},
  doi = {10.1109/twc.2021.3111707},
}
Figure