Rescheduling for Dynamic Traffic in Time-Sensitive Networks

Time-sensitive networking (TSN) has become one of the enhanced Ethernet technologies for real-time applications in future industrial networks. One of the main use cases of TSN is in industrial automation, where it is used to provide deterministic communication among actuators, sensors and controllers in a factory plant. The low latency and low jitter provided by TSN require that time-sensitive traffic be periodic with a constant frame size. However, it remains unclear whether TSN can provide deterministic delay for variable time-sensitive traffic that has variable periods and frame sizes (called dynamic traffic in this paper). A rescheduling method that optimally reallocates resources for the new demands of the dynamic traffic is needed. This paper proposes an Integer Linear Programming (ILP) formulation and a heuristic algorithm to solve the rescheduling problem of a dynamic time-sensitive flow in TSN networks. We compare the proposed algorithm with existing algorithms. The numerical results demonstrate the feasibility of our algorithm in dealing with dynamic time-sensitive traffic and validate the superiority in reducing end-to-end latency and bandwidth waste.