Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Intent-Aware DRL-Based NOMA Uplink Dynamic Scheduler for IIoT

Salwa Mostafa, Mateus P. Mota, Alvaro Valcarce, Mehdi Bennis

TL;DR

The paper tackles IIoT uplink scheduling under URLLC constraints where UEs express high-level intents via a GUI. It introduces a centralized DRL-based dynamic scheduler that uses a graph-based reduction to shrink the large state-action space, enabling efficient learning with PPO and D2RL architectures. The approach achieves higher numbers of completed tasks and improved goodput compared to traditional schemes and contention-based/free baselines, with faster convergence due to the reduction strategy. This work offers a practical, scalable framework for intent-aware MEC-assisted IIoT scheduling in dynamic wireless environments, combining NOMA uplink access with intelligent resource allocation.

Abstract

We investigate the problem of supporting Industrial Internet of Things user equipment (IIoT UEs) with intent (i.e., requested quality of service (QoS)) and random traffic arrival. A deep reinforcement learning (DRL) based centralized dynamic scheduler for time-frequency resources is proposed to learn how to schedule the available communication resources among the IIoT UEs. The proposed scheduler leverages an RL framework to adapt to the dynamic changes in the wireless communication system and traffic arrivals. Moreover, a graph-based reduction scheme is proposed to reduce the state and action space of the RL framework to allow fast convergence and a better learning strategy. Simulation results demonstrate the effectiveness of the proposed intelligent scheduler in guaranteeing the expressed intent of IIoT UEs compared to several traditional scheduling schemes, such as round-robin, semi-static, and heuristic approaches. The proposed scheduler also outperforms the contention-free and contention-based schemes in maximizing the number of successfully computed tasks.

Intent-Aware DRL-Based NOMA Uplink Dynamic Scheduler for IIoT

TL;DR

The paper tackles IIoT uplink scheduling under URLLC constraints where UEs express high-level intents via a GUI. It introduces a centralized DRL-based dynamic scheduler that uses a graph-based reduction to shrink the large state-action space, enabling efficient learning with PPO and D2RL architectures. The approach achieves higher numbers of completed tasks and improved goodput compared to traditional schemes and contention-based/free baselines, with faster convergence due to the reduction strategy. This work offers a practical, scalable framework for intent-aware MEC-assisted IIoT scheduling in dynamic wireless environments, combining NOMA uplink access with intelligent resource allocation.

Abstract

We investigate the problem of supporting Industrial Internet of Things user equipment (IIoT UEs) with intent (i.e., requested quality of service (QoS)) and random traffic arrival. A deep reinforcement learning (DRL) based centralized dynamic scheduler for time-frequency resources is proposed to learn how to schedule the available communication resources among the IIoT UEs. The proposed scheduler leverages an RL framework to adapt to the dynamic changes in the wireless communication system and traffic arrivals. Moreover, a graph-based reduction scheme is proposed to reduce the state and action space of the RL framework to allow fast convergence and a better learning strategy. Simulation results demonstrate the effectiveness of the proposed intelligent scheduler in guaranteeing the expressed intent of IIoT UEs compared to several traditional scheduling schemes, such as round-robin, semi-static, and heuristic approaches. The proposed scheduler also outperforms the contention-free and contention-based schemes in maximizing the number of successfully computed tasks.
Paper Structure (13 sections, 8 equations, 7 figures, 2 tables, 2 algorithms)

This paper contains 13 sections, 8 equations, 7 figures, 2 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Background
  3. Related Work
  4. Contribution
  5. System Model
  6. Network Model
  7. Communication Model
  8. Computation Model
  9. Performance Metrics
  10. Problem Formulation and Proposed Solution
  11. Proximal Policy Optimization (PPO)-based scheduling approach
  12. Simulation Model and Results
  13. Conclusion

Figures (7)

  • Figure 1: Uplink dynamic scheduling.
  • Figure 2: System Model.
  • Figure 3: Normalized successful computed tasks versus training episodes.
  • Figure 4: Normalized failed tasks versus training episodes.
  • Figure 5: Goodput versus training episodes.
  • ...and 2 more figures

Theorems & Definitions (2)

  • Definition 1
  • Example 1