Large-Scale Periodic Scheduling in Time-Sensitive Networks
| Result of project: | Cluster 4.0 – Methodology of System Integration |
| ID project: | EF16_026/0008432 |
| Authors: | Marek Vlk, Kateřina Brejchová, Zdeněk Hanzálek, Siyu Tang |
| Published in: | ELSEVIER: Computers & Operations Research Volume 137, January 2022, 105512 |
| Link: | https://link.springer.com/chapter/10.1007/978-3-662-63519-3_6 |
| DOI: | DOI: 10.1007/978-3-662-63519-3_6 |
Abstract:
The real-time and safety-critical demands posed in a variety of industrial systems can now be handled on the Ethernet architecture by virtue of the IEEE Time-Sensitive Networking task group. In particular, deterministic nature and timeliness guarantees can be achieved by time-triggered communication schedules that are enabled by the IEEE 802.1Qbv standard. However, generating such a schedule is a computationally challenging task due to several harsh constraints that must be satisfied. This problem is usually transformed into some existing formalism, e.g., Satisfiability Modulo Theories, and solved using a general third-party solver, which has the drawback of limited scalability. Except for some one-pass heuristics, such as the earliest deadline first or as-soon-as-possible scheduling, the current literature does not provide any algorithm tailored to this specific problem of scheduling time-triggered traffic in time-sensitive networks. In this paper, we develop an efficient algorithm that does not use any third-party solver and is based on the search in the space of partial schedules while being guided by the discovered conflicts. Our algorithm calculates schedules for problem instances consisting of 2000 network nodes and more than 10 000 flows.
This project has received funding from the Ministry of Education, Youth and Sports, program Operational Programme Research, Development and Education under agreement No. EF16_026/0008432.
