Paper accepted at IFIP Networking 2026

9 April 2026, by Mathias Fischer

Photo: IFIP Networking 2026

We are excited to announce that our paper entitled “On the Impact of gPTP Time Synchronization Attacks on Time-Sensitive Networks” has been accepted for publication at the 25th International Federation for Information Processing (IFIP Networking Conference 2026) in Lugano, Switzerland.

In this work, we analyze protocol-level vulnerabilities in gPTP (IEEE 802.1AS) and their impact on Time-Sensitive Networks (TSN). Through large-scale simulations, we show that even small timing manipulations can propagate through the network, leading to clock divergence, deadline violations, and severe Quality of Service (QoS) degradation.

We are thrilled to present our findings at IFIP Networking in Lugano and look forward to engaging with the research community on the security and resilience of time-sensitive networks.

Paper abstract:

Time synchronization is a fundamental building block for Time-Sensitive Networking (TSN), as it aligns transmission schedules across network nodes. The IEEE 802.1AS specifies the generalized Precision Time Protocol (gPTP), which provides sub-microsecond accuracy but lacks integrity protection, making it vulnerable to message delay, spoofing, and timestamp manipulation. These vulnerabilities pose challenges for TSN scheduling mechanisms, particularly time-triggered approaches like IEEE 802.1Qbv Time-Aware Shaper (TAS), but also rate-based shapers like IEEE 802.1Qav Credit-Based Shaper (CBS).
This paper presents a systematic, end-to-end analysis of how gPTP time synchronization attacks degrade TSN performance. We study three realistic attacker types: a malicious Grandmaster distributing falsified time, cumulative delay injection attacks by compromised boundary clocks, and asymmetric delay attacks that selectively manipulate synchronization messages. We implement these attacks in OMNeT++ and evaluate their impact on CBS in 300 network instances (random, mesh, and scale-free topologies) with mixed-criticality traffic. Results indicate that per-message time offsets up to 5 μs, can accumulate to over 100 μs of clock divergence within one second, causing more than 80% of CBS-scheduled flows to miss deadlines even with a 10% tolerance. Our findings demonstrate the severe vulnerability of TSN scheduling to gPTP attacks across diverse network structures.