[Paper Review] Real-Time Scheduling for 802.1Qbv Time-Sensitive Networking (TSN): A Systematic Review and Experimental Study

Chuanyu Xue, Tianyu Zhang+4 more|arXiv (Cornell University)|May 26, 2023

Network Time Synchronization Technologies73 references4 citations

TL;DR

This paper presents a systematic review and experimental evaluation of 17 state-of-the-art real-time scheduling methods for Time-Aware Shaper (TAS) in Time-Sensitive Networking (TSN), using both high-fidelity simulation and a real TSN testbed. It identifies critical limitations in existing methods, highlights the impact of real-world constraints like PTP message collisions and end-station processing limits, and establishes performance benchmarks for future TSN scheduling research.

ABSTRACT

Time-Sensitive Networking (TSN) has been recognized as one of the key enabling technologies for Industry 4.0 and has been deployed in many mission- and safety-critical applications e.g., automotive and aerospace systems. Given the stringent real-time requirements of these applications, the Time-Aware Shaper (TAS) draws special attention among TSN's many traffic shapers due to its ability to achieve deterministic timing guarantees. Many scheduling methods for TAS shapers have been recently developed that claim to improve system schedulability. However, these scheduling methods have yet to be thoroughly evaluated, especially through experimental comparisons, to provide a systematical understanding of their performance using different evaluation metrics in diverse application scenarios. In this paper, we fill this gap by presenting a systematic review and experimental study on existing TAS-based scheduling methods for TSN. We first categorize the system models employed in these works along with the specific problems they aim to solve, and outline the fundamental considerations in the designs of TAS-based scheduling methods. We then perform an extensive evaluation on 17 representative solutions using both high-fidelity simulations and a real-life TSN testbed, and compare their performance under both synthetic scenarios and real-life industrial use cases. Through these experimental studies, we identify the limitations of individual scheduling methods and highlight several important findings. We expect this work will provide foundational knowledge and performance benchmarks needed for future studies on real-time TSN scheduling.

Motivation & Objective

• To address the lack of systematic, experimental evaluation of TAS-based real-time scheduling methods in TSN networks.
• To identify and analyze the limitations of existing scheduling algorithms under diverse real-world and synthetic scenarios.
• To evaluate the impact of practical constraints—such as PTP message interference, end-station processing limits, and GCL window size bounds—on schedulability and end-to-end delay.
• To provide performance benchmarks and actionable insights for future research on TSN scheduling algorithm design.
• To promote reproducibility by releasing open-source datasets and code for community evaluation of new scheduling methods.

Proposed method

• Categorized 17 representative TAS-based scheduling methods based on their underlying system models, including no-wait, window-based, and JRS models.
• Conducted extensive experiments using a high-fidelity TSN simulator and a real-life TSN testbed with industrial-grade hardware.
• Evaluated performance across multiple metrics: end-to-end delay, jitter, link utilization, schedulability, and GCL length.
• Incorporated real-world constraints into the evaluation, such as co-scheduling of TT traffic and PTP synchronization messages.
• Applied standardized performance metrics, including jitter calculated as the maximum difference in delay between any two frames.
• Validated results on a physical testbed to assess practical feasibility and identify discrepancies between simulation and real-world behavior.

Experimental results

Research questions

• RQ1 How do existing TAS-based scheduling methods perform across different system models and application scenarios in both simulation and real-world environments?
• RQ2 What are the key real-world constraints—such as PTP message interference and end-station processing limitations—that affect the schedulability and end-to-end delay of TT traffic?
• RQ3 Which system model (e.g., no-wait, window-based, JRS) leads to optimal performance in terms of link utilization, GCL length, and schedulability under various workloads?
• RQ4 How do heuristic approaches like the list scheduler compare to model-based optimization methods in large-scale or high-workload scenarios?
• RQ5 To what extent do existing scheduling methods fail to meet real-world deployment requirements due to oversights in modeling practical constraints?

Key findings

• Collisions between TT traffic and PTP synchronization messages can cause significant sync errors, leading to network instability or deadline misses if not properly co-scheduled.
• End stations often impose stricter constraints than TSN bridges due to limited processing capabilities, requiring large inter-frame gaps to preserve packet order—a factor overlooked by most existing methods.
• Only a few scheduling methods account for the GCL window size bound in TSN bridges, making real-world adaptation difficult or impossible when this constraint is ignored.
• Heuristic methods like the list scheduler outperform complex models in high-workload scenarios, suggesting practicality over theoretical optimality in some cases.
• The JRS model leads to lower link utilization and is less efficient than the FR model in large-scale networks, indicating potential inefficiencies in certain modeling approaches.
• The no-wait model is preferred when queue count is limited, as it avoids unnecessary buffering and reduces GCL complexity.