Negotiating strict latency limits for dynamic real-time services in vehicular time-sensitive networks

TL;DR

The paper tackles the problem of guaranteeing hard real-time latency in dynamically changing in-vehicle networks by embedding QoS signaling into automotive service discovery and using a central TSN-SDN controller to compute per-queue idle slopes via a network-calculus-based delay-budget approach. It compares TSN standard worst-case analysis with the delay-budget method, showing that standard methods can fail to bound latency in dynamic IVNs, while the delay-budget approach offers provable guarantees at the cost of higher resource reservation. The authors demonstrate, through synthetic and realistic IVN simulations, that their integrated signaling scheme can configure 450 subscriptions within about 11 ms and maintain end-to-end latencies around 1 ms. This work provides a practical pathway to scalable, deadline-safe dynamic services in future software-defined, service-oriented automotive networks. It also highlights trade-offs between signaling overhead, resource utilization, and latency guarantees, suggesting directions for extending the framework to other TSN mechanisms and service protocols.

Abstract

Future vehicles are expected to dynamically deploy in-vehicle applications within a Service-Oriented Architecture (SOA) while critical services continue to operate under hard real-time constraints. Time-Sensitive Networking (TSN) on the in-vehicle Ethernet layer is dedicated to ensure deterministic communication between critical services; its Credit-Based Shaper (CBS) supports dynamic resource reservations. However, the dynamic nature of service deployment challenges network resource configuration, since any new reservation may change the latency of already validated flows. Standard methods of worst-case latency analysis for CBS have been found incorrect, and current TSN stream reservation procedures lack mechanisms to signal application layer Quality-of-Service (QoS) requirements or verify deadlines. In this paper, we propose and validate a QoS negotiation scheme that interacts with the TSN network controller to reserve resources while ensuring latency bounds. For the first time, this work comparatively evaluates reservation schemes using worst-case analysis and simulations of a realistic In-Vehicle Network (IVN) and demonstrates their impact on QoS guarantees, resource utilization, and setup times. We find that only one reservation scheme utilizing per-queue delay budgets and network calculus provides valid configurations and guarantees acceptable latency bounds throughout the IVN. The proposed service negotiation mechanism efficiently establishes 450 vehicular network reservations in just 11ms.

Figures (14)

• Figure 1: Credit evolution for CBS forwarding.
• Figure 2: Interaction of separate SOA and TSN subscriptions following a two-stage procedure. Subsequent data transmission can start after the SOA subscription. (a) Uses distributed stream reservation in each switch; (b) shows the centralized model of TSN, where signaling of application requirements to the control plane remains unspecified (dashed).
• Figure 3: Network with two flows using CBS and the analytical bound of F1 with and without F2 for various idle slopes at the switch queue. Bandwidth depends on the observed interval, so even when senders comply with their announcements, interference from cross traffic can cause sporadic traffic buildup, leading to higher latency.
• Figure 4: SOA subscription with integrated QoS signaling to the TSSDN control plane using OpenFlow to adapt the network. Subsequent data transmission can start after the SOA subscription is established.
• Figure 5: QoS negotiation sequence for TSSDN within automotive service discovery that integrates bandwidth allocation and deadline validation. Multicast messages, control plane operations, and attached information are highlighted.