Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

EVM Analysis of Distributed Massive MIMO with 1-Bit Radio-Over-Fiber Fronthaul

Anzhong Hu, Lise Aabel, Giuseppe Durisi, Sven Jacobsson, Mikael Coldrey, Christian Fager, Christoph Studer

TL;DR

The paper addresses uplink performance in a distributed massive MIMO system connected by a 1-bit radio-over-fiber fronthaul, where APs forego down-conversion. It leverages Bussgang's decomposition to linearize the 1-bit quantization and analyzes how spatial (AP density) and temporal (oversampling) oversampling interact under a fronthaul capacity constraint to recover transmitted signals, using EVM as the key metric. The authors derive asymptotic characterizations showing the quantization distortion can be whitened into an effective additive noise, identify optimal dithering energy, and compare organized distributed systems to co-located baselines through extensive numerical studies for single- and multi-user scenarios. The results demonstrate that, with sufficient fronthaul capacity, spatial oversampling can outperform temporal oversampling and significantly improve availability and reliability over co-located architectures, while at low fronthaul rates temporal oversampling or fewer APs may be preferable.

Abstract

We analyze the uplink performance of a distributed massive multiple-input multiple-output (MIMO) architecture in which the remotely located access points (APs) are connected to a central processing unit via a fiber-optical fronthaul carrying a dithered and 1-bit quantized version of the received radio-frequency (RF) signal. The innovative feature of the proposed architecture is that no down-conversion is performed at the APs. This eliminates the need to equip the APs with local oscillators, which may be difficult to synchronize. Under the assumption that a constraint is imposed on the amount of data that can be exchanged across the fiber-optical fronthaul, we investigate the tradeoff between spatial oversampling, defined in terms of the total number of APs, and temporal oversampling, defined in terms of the oversampling factor selected at the central processing unit, to facilitate the recovery of the transmitted signal from 1-bit samples of the RF received signal. Using the so-called error-vector magnitude (EVM) as performance metric, we shed light on the optimal design of the dither signal, and quantify, for a given number of APs, the minimum fronthaul rate required for our proposed distributed massive MIMO architecture to outperform a standard co-located massive MIMO architecture in terms of EVM.

EVM Analysis of Distributed Massive MIMO with 1-Bit Radio-Over-Fiber Fronthaul

TL;DR

The paper addresses uplink performance in a distributed massive MIMO system connected by a 1-bit radio-over-fiber fronthaul, where APs forego down-conversion. It leverages Bussgang's decomposition to linearize the 1-bit quantization and analyzes how spatial (AP density) and temporal (oversampling) oversampling interact under a fronthaul capacity constraint to recover transmitted signals, using EVM as the key metric. The authors derive asymptotic characterizations showing the quantization distortion can be whitened into an effective additive noise, identify optimal dithering energy, and compare organized distributed systems to co-located baselines through extensive numerical studies for single- and multi-user scenarios. The results demonstrate that, with sufficient fronthaul capacity, spatial oversampling can outperform temporal oversampling and significantly improve availability and reliability over co-located architectures, while at low fronthaul rates temporal oversampling or fewer APs may be preferable.

Abstract

We analyze the uplink performance of a distributed massive multiple-input multiple-output (MIMO) architecture in which the remotely located access points (APs) are connected to a central processing unit via a fiber-optical fronthaul carrying a dithered and 1-bit quantized version of the received radio-frequency (RF) signal. The innovative feature of the proposed architecture is that no down-conversion is performed at the APs. This eliminates the need to equip the APs with local oscillators, which may be difficult to synchronize. Under the assumption that a constraint is imposed on the amount of data that can be exchanged across the fiber-optical fronthaul, we investigate the tradeoff between spatial oversampling, defined in terms of the total number of APs, and temporal oversampling, defined in terms of the oversampling factor selected at the central processing unit, to facilitate the recovery of the transmitted signal from 1-bit samples of the RF received signal. Using the so-called error-vector magnitude (EVM) as performance metric, we shed light on the optimal design of the dither signal, and quantify, for a given number of APs, the minimum fronthaul rate required for our proposed distributed massive MIMO architecture to outperform a standard co-located massive MIMO architecture in terms of EVM.
Paper Structure (26 sections, 2 theorems, 79 equations, 8 figures)

This paper contains 26 sections, 2 theorems, 79 equations, 8 figures.

Table of Contents

  1. Introduction
  2. System Model
  3. Signal Parameters and Assumptions
  4. Discretized Input-Output Relation
  5. Assumptions
  6. Linear Decomposition Using Bussgang's Theorem
  7. Digital Down-Conversion and Linear Combining
  8. Analysis of the
  9. Bussgang Combiner
  10. Bussgang Combiner
  11. Bussgang Combiner
  12. Asymptotic Characterization of the
  13. The Matrices $\mathbf{G}\xspace\xspace$ and $\mathbf{C}\xspace\xspace_{\hat{\mathbf{e}\xspace\xspace}_{k}}$ in the Frequency-Flat Fading Case, and their Asymptotic Limit
  14. Insights from the Asymptotic Characterization
  15. Numerical Investigations
  16. ...and 11 more sections

Key Result

Lemma 1

The random variable $r_{bb'}[m]$ in eq:ruv-def satisfies the following properties: and, for $b\neq b'$ or $m\neq 0$ Furthermore, for all $b,b'$,

Figures (8)

  • Figure 1: A distributed massive uplink architecture consisting of $B$ connected to a via a fiber-optical fronthaul. The received signal at each AP is filtered, added to a dither signal and compared with a zero threshold. The resulting two-level signal is converted to the optical domain and transmitted over the optical fiber to the , where it is converted back to the electrical domain, oversampled using a $1$-bit , and sent to the for digital down-conversion and spatial processing.
  • Figure 2: The topology of the distributed system for the case $U=1$ and $U=4$ and fixed positions considered in this section. In the two figures, $B=16$.
  • Figure 3: as a function of the dither-to-noise ratio $E_{\text{{}d}}/N_{0}$, for the case of Bussgang / and combiners and $U=1$. The values of $R_{\text{{}fh}}$ are given in Gbit/s. The dot-dashed lines denote the for the case of infinite fronthaul rate, computed using \ref{['eq:evm-mr-as']} and \ref{['eq:evm-mmse-as']}, respectively.
  • Figure 4: as a function of the fronthaul rate $R_{\text{{}fh}}$, for the case of Bussgang / and combiners and $U=1$. The value of $E_{\text{{}d}}$ is optimized for each value of fronthaul rate considered in the figure. The dashed lines denote the for the case of infinite fronthaul rate, computed using \ref{['eq:evm-mr-as']} and \ref{['eq:evm-mmse-as']}, and setting $E_{\text{{}d}}=0$, in accordance to the discussion in Section \ref{['sec:insights']}.
  • Figure 5: as a function of the dither-to-noise ratio $E_{\text{{}d}}/N_{0}$, for the case of Bussgang and combiners and $U=4$. The dot-dashed lines denote the for the case of infinite fronthaul rate, computed by substituting \ref{['eq:limit-matG']} and \ref{['eq:limit-Ce']} into \ref{['eq:evm-zf']} and \ref{['eq:evm-mmse']}.
  • ...and 3 more figures

Theorems & Definitions (2)

  • Lemma 1
  • Theorem 2