Deep-Reinforcement-Learning-Based Scheduler for Time-Aware Shaper in In-Vehicle Networks

Mohammadparsa Karimi; Majid Nabi; Andrew T. Nelson; Kees G.W. Goossens; A.A. (Twan) Basten

doi:10.1109/VTC2025-Spring65109.2025.11174378

Deep-Reinforcement-Learning-Based Scheduler for Time-Aware Shaper in In-Vehicle Networks

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

5 Citations (Scopus)

141 Downloads (Pure)

Abstract

As vehicles develop into software-defined platforms with powerful automated driving capabilities and driver support systems, their in-vehicle networks become significantly more complicated. A key technique for ensuring deterministic, low-latency connectivity for crucial data traffic in such settings is Time-Sensitive Networking (TSN), and specifically the Time-Aware Shaper (TAS). However, current TAS scheduling techniques have difficulty adjusting schedules to dynamically shifting traffic patterns and changing operating conditions. This paper presents an adaptive scheduler using Deep Reinforcement Learning (DRL), which aims to meet strict deadlines, reducing latency and providing near-ideal resource usage. Experimental results for different vehicle scenarios show that our DRL-based scheduler performs better in terms of success rate, low latency, and overall network performance than state-of-the-art heuristic algorithms such as earliest deadline first (EDF) scheduling.

Original language	English
Title of host publication	2025 IEEE 101st Vehicular Technology Conference, VTC2025-Spring
Publisher	Institute of Electrical and Electronics Engineers
Number of pages	7
ISBN (Electronic)	979-8-3315-3147-8
DOIs	https://doi.org/10.1109/VTC2025-Spring65109.2025.11174378
Publication status	Published - 30 Sept 2025
Event	101st Vehicular Technology Conference, VTC2025-Spring - Oslo, Norway Duration: 17 Jun 2025 → 20 Jun 2025

Conference

Conference	101st Vehicular Technology Conference, VTC2025-Spring
Abbreviated title	VTC2025-Spring
Country/Territory	Norway
City	Oslo
Period	17/06/25 → 20/06/25

Funding

This work has received funding from the European Chips Joint Undertaking under Framework Partnership Agreement No 101139789 (HAL4SDV).

Keywords

Adaptive
Deep Reinforcement Learning
In-Vehicle Networking
Scheduling
TAS
TSN

Access to Document

10.1109/VTC2025-Spring65109.2025.11174378

VTC2025-SpringGalley Proof, 1.91 MB
pdfFinal published version, 2.46 MBLicence: TAVERNE

Cite this

@inproceedings{b3f4507db17e4f08b47d825a17c155bb,

title = "Deep-Reinforcement-Learning-Based Scheduler for Time-Aware Shaper in In-Vehicle Networks",

abstract = "As vehicles develop into software-defined platforms with powerful automated driving capabilities and driver support systems, their in-vehicle networks become significantly more complicated. A key technique for ensuring deterministic, low-latency connectivity for crucial data traffic in such settings is Time-Sensitive Networking (TSN), and specifically the Time-Aware Shaper (TAS). However, current TAS scheduling techniques have difficulty adjusting schedules to dynamically shifting traffic patterns and changing operating conditions. This paper presents an adaptive scheduler using Deep Reinforcement Learning (DRL), which aims to meet strict deadlines, reducing latency and providing near-ideal resource usage. Experimental results for different vehicle scenarios show that our DRL-based scheduler performs better in terms of success rate, low latency, and overall network performance than state-of-the-art heuristic algorithms such as earliest deadline first (EDF) scheduling.",

keywords = "Adaptive, Deep Reinforcement Learning, In-Vehicle Networking, Scheduling, TAS, TSN",

author = "Mohammadparsa Karimi and Majid Nabi and Nelson, \{Andrew T.\} and Goossens, \{Kees G.W.\} and Basten, \{A.A. (Twan)\}",

year = "2025",

month = sep,

day = "30",

doi = "10.1109/VTC2025-Spring65109.2025.11174378",

language = "English",

booktitle = "2025 IEEE 101st Vehicular Technology Conference, VTC2025-Spring",

publisher = "Institute of Electrical and Electronics Engineers",

address = "United States",

note = "101st Vehicular Technology Conference, VTC2025-Spring, VTC2025-Spring ; Conference date: 17-06-2025 Through 20-06-2025",

}

Karimi, M , Nabi, M , Nelson, AT , Goossens, KGW & Basten, AA 2025, Deep-Reinforcement-Learning-Based Scheduler for Time-Aware Shaper in In-Vehicle Networks. in 2025 IEEE 101st Vehicular Technology Conference, VTC2025-Spring., 11174378, Institute of Electrical and Electronics Engineers, 101st Vehicular Technology Conference, VTC2025-Spring, Oslo, Norway, 17/06/25. https://doi.org/10.1109/VTC2025-Spring65109.2025.11174378

Deep-Reinforcement-Learning-Based Scheduler for Time-Aware Shaper in In-Vehicle Networks. / Karimi, Mohammadparsa ; Nabi, Majid ; Nelson, Andrew T. et al.
2025 IEEE 101st Vehicular Technology Conference, VTC2025-Spring. Institute of Electrical and Electronics Engineers, 2025. 11174378.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Deep-Reinforcement-Learning-Based Scheduler for Time-Aware Shaper in In-Vehicle Networks

AU - Karimi, Mohammadparsa

AU - Nabi, Majid

AU - Nelson, Andrew T.

AU - Goossens, Kees G.W.

AU - Basten, A.A. (Twan)

PY - 2025/9/30

Y1 - 2025/9/30

N2 - As vehicles develop into software-defined platforms with powerful automated driving capabilities and driver support systems, their in-vehicle networks become significantly more complicated. A key technique for ensuring deterministic, low-latency connectivity for crucial data traffic in such settings is Time-Sensitive Networking (TSN), and specifically the Time-Aware Shaper (TAS). However, current TAS scheduling techniques have difficulty adjusting schedules to dynamically shifting traffic patterns and changing operating conditions. This paper presents an adaptive scheduler using Deep Reinforcement Learning (DRL), which aims to meet strict deadlines, reducing latency and providing near-ideal resource usage. Experimental results for different vehicle scenarios show that our DRL-based scheduler performs better in terms of success rate, low latency, and overall network performance than state-of-the-art heuristic algorithms such as earliest deadline first (EDF) scheduling.

AB - As vehicles develop into software-defined platforms with powerful automated driving capabilities and driver support systems, their in-vehicle networks become significantly more complicated. A key technique for ensuring deterministic, low-latency connectivity for crucial data traffic in such settings is Time-Sensitive Networking (TSN), and specifically the Time-Aware Shaper (TAS). However, current TAS scheduling techniques have difficulty adjusting schedules to dynamically shifting traffic patterns and changing operating conditions. This paper presents an adaptive scheduler using Deep Reinforcement Learning (DRL), which aims to meet strict deadlines, reducing latency and providing near-ideal resource usage. Experimental results for different vehicle scenarios show that our DRL-based scheduler performs better in terms of success rate, low latency, and overall network performance than state-of-the-art heuristic algorithms such as earliest deadline first (EDF) scheduling.

KW - Adaptive

KW - Deep Reinforcement Learning

KW - In-Vehicle Networking

KW - Scheduling

KW - TAS

KW - TSN

UR - https://www.scopus.com/pages/publications/105019041511

U2 - 10.1109/VTC2025-Spring65109.2025.11174378

DO - 10.1109/VTC2025-Spring65109.2025.11174378

M3 - Conference contribution

BT - 2025 IEEE 101st Vehicular Technology Conference, VTC2025-Spring

PB - Institute of Electrical and Electronics Engineers

T2 - 101st Vehicular Technology Conference, VTC2025-Spring

Y2 - 17 June 2025 through 20 June 2025

ER -

Deep-Reinforcement-Learning-Based Scheduler for Time-Aware Shaper in In-Vehicle Networks

Abstract

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this