Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Deep Learning-Based Approach for User Activity Detection with Grant-Free Random Access in Cell-Free Massive MIMO

Ali Elkeshawy, HaÏfa Farès, Amor Nafkha

TL;DR

This work tackles the challenge of reliable user activity detection in grant-free random access for CF-mMIMO-based mMTC. It introduces a Deep Multilayer Perceptron (DMLP) that reframes activity detection as multi-binary label classification, trained on synthetic CF-mMIMO data under non-orthogonal pilot a settings and 3GPP channel models. A clustering-based detection scheme is proposed to reduce computational load while preserving accuracy, and the study analyzes robustness to input perturbations and fixed-point quantization. Results show the DMLP achieves high detection performance (up to 99% accuracy in simulations), with stronger resilience than covariance-based methods across diverse channel conditions and system configurations, highlighting its practical relevance for scalable IoT connectivity in CF-mMIMO networks.

Abstract

Modern wireless networks must reliably support a wide array of connectivity demands, encompassing various user needs across diverse scenarios. Machine-Type Communication (mMTC) is pivotal in these networks, particularly given the challenges posed by massive connectivity and sporadic device activation patterns. Traditional grant-based random access (GB-RA) protocols face limitations due to constrained orthogonal preamble resources. In response, the adoption of grant-free random access (GF-RA) protocols offers a promising solution. This paper explores the application of supervised machine learning models to tackle activity detection issues in scenarios where non-orthogonal preamble design is considered. We introduce a data-driven algorithm specifically designed for user activity detection in Cell-Free Massive Multiple-Input Multiple-Output (CF-mMIMO) networks operating under GF-RA protocols. Additionally, this study presents a novel clustering strategy that simplifies and enhances activity detection accuracy, assesses the resilience of the algorithm to input perturbations, and investigates the effects of adopting floating-to-fixed-point conversion on algorithm performance. Simulations conducted adhere to 3GPP standards, ensuring accurate channel modeling, and employ a deep learning approach to boost the detection capabilities of mMTC GF-RA devices. The results are compelling: the algorithm achieves an exceptional 99\% accuracy rate, confirming its efficacy in real-world applications.

Deep Learning-Based Approach for User Activity Detection with Grant-Free Random Access in Cell-Free Massive MIMO

TL;DR

This work tackles the challenge of reliable user activity detection in grant-free random access for CF-mMIMO-based mMTC. It introduces a Deep Multilayer Perceptron (DMLP) that reframes activity detection as multi-binary label classification, trained on synthetic CF-mMIMO data under non-orthogonal pilot a settings and 3GPP channel models. A clustering-based detection scheme is proposed to reduce computational load while preserving accuracy, and the study analyzes robustness to input perturbations and fixed-point quantization. Results show the DMLP achieves high detection performance (up to 99% accuracy in simulations), with stronger resilience than covariance-based methods across diverse channel conditions and system configurations, highlighting its practical relevance for scalable IoT connectivity in CF-mMIMO networks.

Abstract

Modern wireless networks must reliably support a wide array of connectivity demands, encompassing various user needs across diverse scenarios. Machine-Type Communication (mMTC) is pivotal in these networks, particularly given the challenges posed by massive connectivity and sporadic device activation patterns. Traditional grant-based random access (GB-RA) protocols face limitations due to constrained orthogonal preamble resources. In response, the adoption of grant-free random access (GF-RA) protocols offers a promising solution. This paper explores the application of supervised machine learning models to tackle activity detection issues in scenarios where non-orthogonal preamble design is considered. We introduce a data-driven algorithm specifically designed for user activity detection in Cell-Free Massive Multiple-Input Multiple-Output (CF-mMIMO) networks operating under GF-RA protocols. Additionally, this study presents a novel clustering strategy that simplifies and enhances activity detection accuracy, assesses the resilience of the algorithm to input perturbations, and investigates the effects of adopting floating-to-fixed-point conversion on algorithm performance. Simulations conducted adhere to 3GPP standards, ensuring accurate channel modeling, and employ a deep learning approach to boost the detection capabilities of mMTC GF-RA devices. The results are compelling: the algorithm achieves an exceptional 99\% accuracy rate, confirming its efficacy in real-world applications.
Paper Structure (21 sections, 20 equations, 14 figures, 1 table)

This paper contains 21 sections, 20 equations, 14 figures, 1 table.

Table of Contents

  1. Introduction
  2. Literature Review
  3. Contributions
  4. System Model
  5. Deep Learning Algorithm
  6. Data Generation
  7. Training Phase
  8. Deep Multilayer Perceptron (DMLP)
  9. Clustering-Based Activity Detection
  10. Comparative Analysis of Scenarios and Algorithms Applicability
  11. Algorithm Applicability
  12. DMLP Performance Summary
  13. Numerical Results
  14. DMLP Parameters Tuning
  15. SNR Analysis Results
  16. ...and 6 more sections

Figures (14)

  • Figure 1: Cell-Free network model for mMTC.
  • Figure 2: Distribution of users and access points within the network.
  • Figure 3: Architecture of the DMLP algorithm proposed for UAD.
  • Figure 4: DMLP algorithm performance in scenario II from ganesan2020clustering.
  • Figure 5: Mathematical approach performance in scenario I.
  • ...and 9 more figures