For some industrial use cases, the end user might be interested in the performance of a logical link in one direction (only uplink or only downlink), rather than the round-trip performance. This is the case for many industrial users for which end-to-end performance for automation functions is crucial. Measuring the one-way end-to-end transmission time is different from measuring the round-trip performance and dividing by two, since there are usually differences in the uplink (UL) and downlink (DL) performance and in the processing time in the server that reflects the message. A measurement system can therefore be used to measure performance in the UL or DL direction only.
Two measurement devices would be used in this case: one connected to the device in the field in which the use case would be implemented, and another connected to the managed switch of the 5G core or even further along, depending on the architecture of the use case connection. Messages are sent between the two devices in one direction or concurrently in both directions, but only measured in the one-way direction. The measurement devices can be synchronized to enable measurement of the transmission time between them and other performance-related parameters of the communication system. In many cases, it is important to properly analyze and take into account the performance of every sent and received message instead of just statistical averages. Highly accurate (to within one microsecond) precision and determinism of measurements are even more important; this is very difficult for normal CPUs to match.
Alternatively, a UDP stream test can be used for one-way measurements. The UDP stream test sends a predefined packet stream in UL or DL and analyzes how it is received. A server streams packets in accordance with defined parameters to a measurement device (or vice versa). Independent uplink and downlink traffic with more specific traffic patterns can be generated. A packet stream isn’t only defined by the bitrate, but also by the packet sizes and frequency and whether the packets are delivered in bursts or singly. Depending on these settings, the packet loss, latency, and jitter in inter-packet arrival times can also vary at a constant bitrate.
The figure below shows an example of this logical link measurement between two local application functions, with the source measurement device sending emulated application messages to the target measurement device.
One important aspect of the logical link measurement is that it isn’t restricted to measuring the performance of the wireless communication system itself. If the use case isn’t connected to the user equipment (UE) directly but instead via one or more switches or other devices, this also reveals the effect of interplay and cooperation among the wireless communication system and other nodes along the communication path. This means that the system is measured from the application’s perspective while taking all media transitions and device behaviors into account.
