Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Joint Link Adaptation and Device Scheduling Approach for URLLC Industrial IoT Network: A DRL-based Method with Bayesian Optimization

Wei Gao, Paul Zheng, Peng Wu, Yulin Hu, Anke Schmeink

TL;DR

This work tackles joint link adaptation and device scheduling in a dynamic IIoT URLLC downlink with imperfect CSI. It integrates Bayesian optimization with TD3, augmented by an OLLA execution phase and a GEXP-BO training module to accelerate convergence and stabilize learning under ACK/NACK imbalances. The approach demonstrates faster convergence, higher sum-rate performance, and improved reliability compared to baselines, even when MCS must be chosen from discrete 3GPP indices. The framework offers robust and scalable performance for URLLC IIoT deployments, with strong practical implications for real-time, reliable command delivery in dense device networks.

Abstract

In this article, we consider an industrial internet of things (IIoT) network supporting multi-device dynamic ultra-reliable low-latency communication (URLLC) while the channel state information (CSI) is imperfect. A joint link adaptation (LA) and device scheduling (including the order) design is provided, aiming at maximizing the total transmission rate under strict block error rate (BLER) constraints. In particular, a Bayesian optimization (BO) driven Twin Delayed Deep Deterministic Policy Gradient (TD3) method is proposed, which determines the device served order sequence and the corresponding modulation and coding scheme (MCS) adaptively based on the imperfect CSI. Note that the imperfection of CSI, error sample imbalance in URLLC networks, as well as the parameter sensitivity nature of the TD3 algorithm likely diminish the algorithm's convergence speed and reliability. To address such an issue, we proposed a BO based training mechanism for the convergence speed improvement, which provides a more reliable learning direction and sample selection method to track the imbalance sample problem. Via extensive simulations, we show that the proposed algorithm achieves faster convergence and higher sum-rate performance compared to existing solutions.

Joint Link Adaptation and Device Scheduling Approach for URLLC Industrial IoT Network: A DRL-based Method with Bayesian Optimization

TL;DR

This work tackles joint link adaptation and device scheduling in a dynamic IIoT URLLC downlink with imperfect CSI. It integrates Bayesian optimization with TD3, augmented by an OLLA execution phase and a GEXP-BO training module to accelerate convergence and stabilize learning under ACK/NACK imbalances. The approach demonstrates faster convergence, higher sum-rate performance, and improved reliability compared to baselines, even when MCS must be chosen from discrete 3GPP indices. The framework offers robust and scalable performance for URLLC IIoT deployments, with strong practical implications for real-time, reliable command delivery in dense device networks.

Abstract

In this article, we consider an industrial internet of things (IIoT) network supporting multi-device dynamic ultra-reliable low-latency communication (URLLC) while the channel state information (CSI) is imperfect. A joint link adaptation (LA) and device scheduling (including the order) design is provided, aiming at maximizing the total transmission rate under strict block error rate (BLER) constraints. In particular, a Bayesian optimization (BO) driven Twin Delayed Deep Deterministic Policy Gradient (TD3) method is proposed, which determines the device served order sequence and the corresponding modulation and coding scheme (MCS) adaptively based on the imperfect CSI. Note that the imperfection of CSI, error sample imbalance in URLLC networks, as well as the parameter sensitivity nature of the TD3 algorithm likely diminish the algorithm's convergence speed and reliability. To address such an issue, we proposed a BO based training mechanism for the convergence speed improvement, which provides a more reliable learning direction and sample selection method to track the imbalance sample problem. Via extensive simulations, we show that the proposed algorithm achieves faster convergence and higher sum-rate performance compared to existing solutions.
Paper Structure (26 sections, 24 equations, 10 figures, 3 tables, 2 algorithms)

This paper contains 26 sections, 24 equations, 10 figures, 3 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Work
  3. Optimization based link adaptation and device scheduling
  4. Machine Learning based link adaptation and device scheduling
  5. DRL based link adaptation and device scheduling
  6. Systems Model
  7. Control Parameter Phase
  8. Transmission and Feedback Phase
  9. Problem Statement
  10. BO Assisted Deep Reinforcement Learning Framework
  11. Preliminary of Conventional DRL Method
  12. OLLA Assisted Execution Phase
  13. BO Based Training Phase
  14. GEXP-assisted BO Based Joint RD and MCS Selection Algorithm
  15. Sample Selection Method
  16. ...and 11 more sections

Figures (10)

  • Figure 1: The TDMA based transmission frame structure for the IIoT networks.
  • Figure 2: Transmission and Feedback Phase Process Diagram.
  • Figure 3: The Flowchart of BO-TD3 Algorithm.
  • Figure 4: Sum rate performance of different schemes over training epoch.
  • Figure 5: Test performance of different schemes versus communication round.
  • ...and 5 more figures