CNN-Enabled Scheduling for Probabilistic Real-Time Guarantees in Industrial URLLC

TL;DR

The paper tackles per-packet real-time guarantees in industrial URLLC across multi-cell, multi-channel wireless networks. It advances Local Deadline Partition by integrating CNN-based priority prediction with graph-coloring for interference-aware RB allocation, trained offline. The approach yields substantial SINR gains, 100% schedulability, and high reliability across network scales, demonstrating strong potential for real-time, scalable scheduling in industrial settings. These results highlight the practical impact of combining CNN priors with graph-based resource allocation for complex URLLC scenarios.

Abstract

Ensuring packet-level communication quality is vital for ultra-reliable, low-latency communications (URLLC) in large-scale industrial wireless networks. We enhance the Local Deadline Partition (LDP) algorithm by introducing a CNN-based dynamic priority prediction mechanism for improved interference coordination in multi-cell, multi-channel networks. Unlike LDP's static priorities, our approach uses a Convolutional Neural Network and graph coloring to adaptively assign link priorities based on real-time traffic, transmission opportunities, and network conditions. Assuming that first training phase is performed offline, our approach introduced minimal overhead, while enabling more efficient resource allocation, boosting network capacity, SINR, and schedulability. Simulation results show SINR gains of up to 113\%, 94\%, and 49\% over LDP across three network configurations, highlighting its effectiveness for complex URLLC scenarios.

Figures (4)

• Figure 1: Conflict graph with 6 nodes, where each node represents a communication link, and edges indicate interference relationships.
• Figure 2: Architecture of CNN Models.
• Figure 3: Training Losses of our proposed CNN Models: priority prediction model and Resource Block Selection model.
• Figure 4: Interference Effect on Receiver-Side SINR.