User Scheduling for Millimeter Wave Hybrid Beamforming Systems with Low-Resolution ADCs

TL;DR

This paper develops new user scheduling criteria for mmWave systems with hybrid beamforming and low-resolution ADCs, optimizing achievable rates by considering channel structure, and proposes efficient algorithms validated by simulations.

Contribution

It introduces novel scheduling criteria accounting for beamspace channel structure and proposes algorithms tailored for low-resolution ADC mmWave systems.

Findings

Scheduling based on channel paths and AoAs improves rates.

Chordal distance scheduling exploits AoA knowledge effectively.

Quantization error impacts sum rate, mitigated by phase offset considerations.

Abstract

We investigate uplink user scheduling for millimeter wave (mmWave) hybrid analog/digital beamforming systems with low-resolution analog-to-digital converters (ADCs). Deriving new scheduling criteria for the mmWave systems, we show that the channel structure in the beamspace, in addition to the channel magnitude and orthogonality, plays a key role in maximizing the achievable rates of scheduled users due to quantization error. The criteria show that to maximize the achievable rate for a given channel gain, the channels of the scheduled users need to have (1) as many propagation paths as possible with unique angle-of-arrivals (AoAs) and (2) even power distribution in the beamspace. Leveraging the derived criteria, we propose an efficient scheduling algorithm for mmWave zero-forcing receivers with low-resolution ADCs. We further propose a chordal distance-based scheduling algorithm that exploits only the AoA knowledge and analyze the performance by deriving ergodic rates in closed form. Based on the derived rates, we show that the beamspace channel leakage resulting from phase offsets between AoAs and quantized angles of analog combiners can lead to sum rate gain by reducing quantization error compared to the channel without leakage. Simulation results validate the sum rate performance of the proposed algorithms and derived ergodic rates expressions.