A Study on Determination of an Appropriate GCL of Time-Aware Shaper in Ethernet-Based Industrial Networks

In Ethernet-based industrial networks, adopting QoS control defined by IEEE 802.1TSN (Time-Sensitive Networking) is being considered to transmit time-sensitive traffic. Although some QoS controls are defined in IEEE 802.1TSN, Time-Aware Shaper (TAS) can control QoS with the highest precision and is expected to be widely used in industrial use. Since TAS handles the opening and closing of queue gates at output ports for each priority based on a GCL (Gating Control List), setting the appropriate GCL to meet user requirements is necessary. However, it is difficult to find an optimal GCL because it is NP-hard; a method for determining such a GCL needs to be studied. This paper proposes a method to find an appropriate GCL using a genetic algorithm, one of the numerical optimization methods; it evaluates the method's effectiveness by simulation-based experiments.