[Paper Review] A Low-Resolution ADC Module Assisted Hybrid Beamforming Architecture for mmWave Communications

Jie Yang, Xi Yang+3 more|arXiv (Cornell University)|Mar 26, 2018

Millimeter-Wave Propagation and Modeling7 references3 citations

TL;DR

This paper proposes a low-resolution ADC module-assisted hybrid beamforming architecture for mmWave communications that drastically reduces beam training time and complexity while maintaining high data transmission performance. By switching between a low-resolution ADC and hybrid beamforming modules during training, and using a two-phase compressed sensing-based beam training method, the system achieves accurate beam alignment in only $L+1$ time slots, where $L$ is the number of propagation paths.

ABSTRACT

We propose a low-resolution analog-to-digital converter (ADC) module assisted hybrid beamforming architecture for millimeter-wave (mmWave) communications. We prove that the proposed low-cost and flexible architecture can reduce the beam training time and complexity dramatically without degradation in the data transmission performance. In addition, we design a fast beam training method which is suitable for the proposed system architecture. The proposed beam training method requires only L + 1 (where L is the number of paths) time slots which is smaller compared to the state-of-the-art.

Motivation & Objective

• To address the high hardware complexity and long beam training times in mmWave hybrid beamforming systems.
• To reduce power consumption and computational load at the receiver by employing low-resolution ADCs.
• To design a fast, efficient beam training method compatible with low-resolution ADCs.
• To maintain high spectral efficiency and data transmission performance despite reduced ADC resolution.
• To enable hardware-realizable, low-cost mmWave transceivers with minimal training overhead.

Proposed method

• The system uses an electronic switch to alternate between a low-resolution ADC module and a hybrid beamforming module at both base station and mobile station.
• During beam training, the transmitter uses the hybrid beamforming module while the receiver uses the low-resolution ADC module, ensuring no increase in total power consumption.
• A two-phase beam training method is proposed: Phase 1 performs all-directions transmitting to estimate all possible AoAs, and Phase 2 matches transmitted and received beams via compressed sensing.
• The method leverages compressed sensing to reconstruct channel state information from $G_t imes L$ measurements using 1-bit or 2-bit quantized feedback.
• The channel is modeled as a sparse response in angular domain, with $L$ paths and directional beams represented via UPA response vectors.
• The beam alignment is achieved by identifying the strongest entries in the reconstructed channel matrix, using orthogonal matching pursuit for signal recovery.

Experimental results

Research questions

• RQ1 How can low-resolution ADCs be integrated into a hybrid beamforming architecture to reduce hardware complexity without sacrificing performance?
• RQ2 What beam training method enables fast and accurate beam alignment under low-resolution ADC constraints?
• RQ3 Can the proposed architecture achieve beam training in fewer time slots than conventional methods?
• RQ4 What is the impact of 1-bit vs. 2-bit ADC resolution on beam training success rate and system performance?
• RQ5 How does increasing the number of grid points affect the accuracy and reliability of beam alignment?

Key findings

• The proposed beam training method requires only $L+1$ time slots, significantly reducing training overhead compared to conventional methods.
• With 2-bit quantization and $L=1$, the beam training success rate reaches 100% at $-10$ dB SNR.
• With 1-bit quantization and $L=2$, the success rate reaches 100% at 0 dB SNR.
• Increasing the number of grid points ($G_t, G_r = 32$) improves the success rate compared to $G_t, G_r = 16$, confirming scalability.
• The proposed method reduces beam training time from $20T_{\text{Slot}}$ (basic method) to $3T_{\text{Slot}}$ under typical settings, demonstrating substantial efficiency gains.
• The system maintains high data transmission performance despite low-resolution ADCs, proving the feasibility of low-cost mmWave transceivers.