mmWave Massive MIMO with Simple RF and Appropriate DSP

TL;DR

This paper explores cost-effective mmWave massive MIMO solutions using simple RF hardware and advanced DSP techniques to enable practical deployment in next-generation wireless systems.

Contribution

It proposes methods like low-resolution ADCs, wireless LO distribution, and spatial multiplexing to reduce cost and power consumption in mmWave massive MIMO systems.

Findings

Low-resolution ADCs can significantly cut costs.

Wireless LO distribution simplifies hardware architecture.

Spatial multiplexing enhances data rates with fewer resources.

Abstract

There is considerable interest in the combined use of millimeter-wave (mmwave) frequencies and arrays of massive numbers of antennas (massive MIMO) for next-generation wireless communications systems. A symbiotic relationship exists between these two factors: mmwave frequencies allow for densely packed antenna arrays, and hence massive MIMO can be achieved with a small form factor; low per-antenna SNR and shadowing can be overcome with a large array gain; steering narrow beams or nulls with a large array is a good match for the line-of-sight (LOS) or near-LOS mmwave propagation environments, etc.. However, the cost and power consumption for standard implementations of massive MIMO arrays at mmwave frequencies is a significant drawback to rapid adoption and deployment. In this paper, we examine a number of possible approaches to reduce cost and power at both the basestation and user terminal, making up for it with signal processing and additional (cheap) antennas. These approaches include lowresolution Analog-to-Digital Converters (ADCs), wireless local oscillator distribution networks, spatial multiplexing and multistreaming instead of higher-order modulation etc.. We will examine the potential of these approaches in making mmwave massive MIMO a reality and discuss the requirements in terms of digital signal processing (DSP).