Codebook Based Hybrid Precoding for Millimeter Wave Multiuser Systems

Shiwen He, Jiaheng Wang, Yongming Huang, Björn E. Ottersten, Wei HongIEEE TSP 2017 · DOI · OpenReview

TL;DR

This paper tackles the challenge of efficient wireless communication in millimeter-wave (mmWave) systems by proposing a novel codebook-based hybrid precoding technique. This method significantly reduces the complexity and cost of mmWave systems by optimizing both analog RF and digital baseband precoders, achieving high spectral and energy efficiency while overcoming hardware limitations.

What Problem Does This Paper Address?

Millimeter-wave (mmWave) communication offers a promising solution for increasing wireless network capacity due to its vast available bandwidth. However, mmWave signals suffer from severe path loss, requiring sophisticated antenna systems to compensate. Traditional fully digital precoding, where each antenna has its own RF chain, becomes prohibitively expensive and power-hungry due to the large number of antennas typically used in mmWave systems. Hybrid precoding, which combines analog RF and digital baseband processing, offers a more practical solution. The challenge lies in efficiently designing these hybrid precoders, especially in multiuser scenarios with limited RF chains, while simultaneously optimizing for both data rate (spectral efficiency) and power consumption (energy efficiency). Existing methods often assume perfect channel state information (CSI), which is difficult to obtain in practice, or struggle with the complexity of joint optimization.

Methods

The authors propose a codebook-based hybrid precoding design that simplifies the optimization problem. Instead of optimizing the RF precoder directly, they leverage a predefined codebook of RF beamforming vectors. This significantly reduces the search space for the RF precoder. To address the difficulty of obtaining accurate CSI, they employ a beam sweep procedure. This procedure, combined with the directional nature of mmWave channels, allows for efficient estimation of the effective channel in the beam domain, reducing the feedback overhead.

The core of their approach involves transforming the original problem into a Joint Codeword Selection and Precoder Design (JWSPD) problem. This involves selecting the optimal RF codewords from the codebook and designing the baseband precoder simultaneously. They develop efficient algorithms to solve these JWSPD problems for two objectives: maximizing the system sum rate (SRmax) and maximizing the system energy efficiency (EEmax). For the SRmax problem, they use a combination of convex optimization techniques and a sparse regularization approach to handle the non-convex constraints. For the EEmax problem, which is more complex due to its fractional objective and power consumption terms, they also employ convex approximations and iterative optimization methods.

Datasets

Not discussed in the paper. The paper focuses on theoretical development and simulation-based evaluation using synthetic channel models.

Key Insights and Ideas

  • Codebook-based RF precoding simplifies complexity: By restricting RF precoder choices to a predefined codebook, the optimization problem becomes significantly more tractable.
  • Beam sweep for efficient CSI acquisition: Utilizing the beam-domain sparsity of mmWave channels allows for reduced CSI feedback, making the system more practical.
  • JWSPD framework unifies optimization: The problem of selecting RF codewords and designing baseband precoders is effectively combined into a single optimization framework.
  • Non-monotonicity of EE with respect to sparsity: The energy efficiency does not always increase with fewer RF chains; there's a sweet spot, necessitating a dynamic search for the optimal sparse parameter.
  • Trade-off between SR and EE: The paper implicitly highlights the trade-off between maximizing data rate and minimizing power consumption, offering solutions for both objectives.

Key Results

The proposed codebook-based hybrid precoding method demonstrates significant performance improvements compared to existing hybrid precoding schemes and approaches the performance of fully digital precoding. Specifically:

  • Spectral Efficiency: The proposed SRmax hybrid precoding achieves the highest sum rate among the compared methods, outperforming fully analog beamforming and other hybrid approaches. For example, in Fig. 5, it shows a higher sum rate than "Hybrid Precoding Scheme (PZF)" and "Two-stage Hybrid Precoding" across various SNR levels.
  • Energy Efficiency: The proposed EEmax hybrid precoding also shows competitive performance, especially in scenarios with limited RF chains, offering a better EE compared to fully digital precoding. Fig. 8 and 9 illustrate this, showing that the proposed EEmax precoder (using DFT codebook) can achieve better EE than fully digital precoding.
  • Reduced Hardware Cost: The hybrid approach inherently leads to lower hardware costs due to the reduced number of RF chains compared to fully digital systems, which is crucial for mmWave systems.
  • Effective RF Chain Selection: The algorithms effectively select a subset of RF chains (codewords) to meet performance targets, balancing efficiency and complexity. Table I provides examples of how different values of the sparsity parameter $\lambda$ lead to different sets of selected RF chains and corresponding EE values.

Limitations and Future Work

  • Perfect CSI assumption for users: While the paper addresses CSI acquisition at the BS, it implicitly assumes users can provide feedback for their effective channels.
  • Idealized channel model: The channel model is based on the extended Saleh-Valenzuela model, which might not capture all real-world mmWave propagation complexities.
  • Computational complexity: While the proposed methods are more efficient than fully digital precoding, the optimization problems can still be computationally intensive, especially for large systems.
  • Future Work: The authors suggest that the proposed framework can be extended to include power consumption at the receivers. Further research could explore more advanced channel estimation techniques and adaptive codebook designs. The paper also hints at the potential for extending the framework to other performance measures and system configurations.

Summary updated

Abstract

In millimeter wave (mmWave) systems, antenna architecture limitations make it difficult to apply conventional fully digital precoding techniques but call for low cost analog radio-frequency (RF) and digital baseband hybrid precoding methods. This paper investigates joint RF-baseband hybrid precoding for the downlink of multiuser multi-antenna mmWave systems with a limited number of RF chains. Two performance measures, maximizing the spectral efficiency and the energy efficiency of the system, are considered. We propose a codebook based RF precoding design and obtain the channel state information via a beam sweep procedure. Via the codebook based design, the original system is transformed into a virtual multiuser downlink system with the RF chain constraint. Consequently, we are able to simplify the complicated hybrid precoding optimization problems to joint codeword selection and precoder design (JWSPD) problems. Then, we propose efficient methods to address the JWSPD problems and jointly optimize the RF and baseband precoders under the two performance measures. Finally, extensive numerical results are provided to validate the effectiveness of the proposed hybrid precoders.
BibTeX
@article{he2017codebook,
  author = {Shiwen He and Jiaheng Wang and Yongming Huang and Björn E. Ottersten and Wei Hong},
  title = {{Codebook Based Hybrid Precoding for Millimeter Wave Multiuser Systems}},
  journal = {IEEE Trans. Signal Process.},
  year = {2017},
  doi = {10.1109/tsp.2017.2723353},
}
Figure