PSFP in TSN Networks: Insights into Some Practical Limitations

Per-Stream Filtering and Policing (PSFP), standardized in IEEE 802.1Qci, is a mechanism for providing fault containment in Time-Sensitive Networking (TSN) networks. This paper examines limitations of PSFP, showing that the fault-containment can be insufficient. In particular, the Flow Meter inside PSFP measures traffic in Service Data Unit (SDU) bytes – i.e., from the MAC destination address through the Frame Check Sequence. However, common Ethernet shapers such as the Credit-Based Shaper (CBS) regulate traffic based on the full 'on-wire' packet length, which includes the SDU plus the 8-byte preamble and the 12-byte inter-frame gap. This results in a 20-byte per-frame gap that increases admissible rates: with minimum-size packets, a talker can exceed its contractual bandwidth by up to 30%. In addition to the contract not being enforced, an independent stream might be penalized due to a queue build up. Through simulations with RTaW-Pegase software and hardware-in-the-loop experiments on a TSN testbed, we quantify these effects and evaluate configuration-level mitigations. We then discuss possible evolutions of the standard, including overhead-aware byte counting, which could address these gaps.