One of the most important features of an AVB/TSN network is that reserved streams pass through with a low and bounded latency. In legacy Ethernet, the traffic flow of data is indeterminate, intervening traffic can delay stream data. Because of this uncertainty on receiving a stream packet, the receivers in legacy systems typically employ large buffers so as not to underflow, which would result in an audible click on an audio stream or lose critical control information in a control stream.
AVB/TSN not only ensures the arrival of time-sensitive streams, it ensures when they will arrive. The FQTSS (Forwarding and Queuing of Time-Sensitive Streams) standard prioritizes AVB traffic ahead of legacy best-effort packets. AVB frames are forwarded with precedence over Best Effort traffic (i.e., reserved AVB stream traffic traversing an AVB bridge has forwarding precedence over non-reserved traffic) and will be subjected to traffic shaping rules.
The traffic shaping rules for bridges require that frames be distributed very evenly in time. This means that all the AVB/TSN traffic being transmitted out a particular port is distributed evenly in time. This has the effect of smoothing out the delivery times – preventing “bunching” of frames.
By controlling the queuing and data flow, an AVB/TSN system has the lowest possible latency. That latency is controlled and bounded by the rules of FQTSS. If a system has fewer switch hops then the designer can reduce the latency even further and be assured that the traffic will arrive correctly.