Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Type-Based Unsourced Multiple Access over Fading Channels with Cell-Free Massive MIMO

Kaan Okumus, Khac-Hoang Ngo, Giuseppe Durisi, Erik G. Ström

TL;DR

This paper addresses type-based unsourced multiple access (TUMA) over fading channels in a cell-free massive MIMO setting. It proposes CSI-free, multisource AMP-based decoders in centralized and distributed configurations, using location-based codeword partition and Monte Carlo approximations to estimate multiplicities and the type without channel state information. The central contribution is a practical, scalable framework that outperforms CSI-reliant baselines like AMP-DA under imperfect CSI, with the distributed variant offering substantial computational savings for large CF-MIMO deployments. The results demonstrate robust type estimation and scalable operation for massive machine-type communications in realistic wireless networks.

Abstract

Type-based unsourced multiple access (TUMA) is a recently proposed framework for type-based estimation in massive uncoordinated access networks. We extend the existing design of TUMA, developed for an additive white Gaussian channel, to a more realistic environment with fading and multiple antennas. Specifically, we consider a cell-free massive multiple-input multiple-output system and exploit spatial diversity to estimate the set of transmitted messages and the number of users transmitting each message. Our solution relies on a location-based codeword partition and on the use at the receiver of a multisource approximate message passing algorithm in both centralized and distributed implementations. The proposed TUMA framework results in a robust and scalable architecture for massive machine-type communications.

Type-Based Unsourced Multiple Access over Fading Channels with Cell-Free Massive MIMO

TL;DR

This paper addresses type-based unsourced multiple access (TUMA) over fading channels in a cell-free massive MIMO setting. It proposes CSI-free, multisource AMP-based decoders in centralized and distributed configurations, using location-based codeword partition and Monte Carlo approximations to estimate multiplicities and the type without channel state information. The central contribution is a practical, scalable framework that outperforms CSI-reliant baselines like AMP-DA under imperfect CSI, with the distributed variant offering substantial computational savings for large CF-MIMO deployments. The results demonstrate robust type estimation and scalable operation for massive machine-type communications in realistic wireless networks.

Abstract

Type-based unsourced multiple access (TUMA) is a recently proposed framework for type-based estimation in massive uncoordinated access networks. We extend the existing design of TUMA, developed for an additive white Gaussian channel, to a more realistic environment with fading and multiple antennas. Specifically, we consider a cell-free massive multiple-input multiple-output system and exploit spatial diversity to estimate the set of transmitted messages and the number of users transmitting each message. Our solution relies on a location-based codeword partition and on the use at the receiver of a multisource approximate message passing algorithm in both centralized and distributed implementations. The proposed TUMA framework results in a robust and scalable architecture for massive machine-type communications.
Paper Structure (32 sections, 45 equations, 5 figures, 1 algorithm)

This paper contains 32 sections, 45 equations, 5 figures, 1 algorithm.

Table of Contents

  1. Introduction
  2. System Model
  3. Messages and Encoder
  4. Multiplicity and Type
  5. Channel Model
  6. Decoder
  7. Proposed Decoder
  8. Centralized Decoder
  9. Multisource AMP
  10. Prior Selection
  11. Denoiser
  12. Onsager Correction
  13. Multiplicity Estimation
  14. Approximation Methods for Efficient Implementation
  15. Distributed Decoder
  16. ...and 17 more sections

Figures (5)

  • Figure 1: An example topology of the proposed TUMA framework within a CF massive MIMO network over $\mathrm{U}\xspace=9$ zones.
  • Figure 2: Block diagram of the proposed TUMA framework with fading channel in a CF system.
  • Figure 3: The average total variation $\overline{\mathbb{T}\xspace\mathbb{V}\xspace}$ vs. the number of bits per zone $\log_2 \mathrm{M}\xspace$ for $\text{SNR}_{\text{rx}} = \qty{-30}{dB}$ with centralized decoder.
  • Figure 4: The average total variation $\overline{\mathbb{T}\xspace\mathbb{V}\xspace}$ vs. received signal to noise ratio $\text{SNR}_{\text{rx}}$ for $\mathrm{N}\xspace=1024$ and ${\mathrm{M}\xspace} = 2^8$.
  • Figure 5: The average total variation $\overline{\mathbb{T}\xspace\mathbb{V}\xspace}$ vs. maximum phase shift $\phi_{\text{max}}$ for imperfect CSI with ${\mathrm{M}\xspace}=2^8$ and $\text{SNR}_{\text{rx}} = \qty{10}{dB}$.