Delay Asymmetry DetectionMeasurement

1

(% class="row" %)

2

(((

3

(% class="col-xs-12 col-sm-7 test-report-content" %)

(((

----

Controlling and handling asymmetric delay on links that carry PTP messages is crucial for network time synchronization and presents significant challenges.

8

We performed this test in four different combinations. The setup included one Grandmaster (either real or emulated) with an impaired link to Boundary Clock-1.

9

When we used an emulated T-GM, like the Calnex Paragon-Neo PAM4 or Keysight Time Sync Analyzer, the impairment came directly from the emulated T-GM.

10

If we used a real T-GM, specifically the Microchip TimeProvider® 4500, we used an external impairment device, the Calnex SNE Ignite.

11

Initially, we connected T-BC-1 directly to a GNSS reference and connected Boundary Clock-2 (T-BC-2) to a time error measurement device (either the Calnex Paragon-Neo PAM4 or Keysight Time Sync Analyzer).

12

We utilized the PTP profile ITU-T G.8275.2 for three of the four combinations throughout the entire chain.

13

In the fourth combination, we implemented the interworking function (IWF) to transition from the PTP profile ITU-T G.8275.2 to ITU-T G.8275.1 between T-BC-1 (Ciena 5134) and T-BC-2 (Juniper PTX10002-36QDD).

14

Finally, we connected T-BC-2 to the time error measurement device over an 800G link.

15

16

[[Figure 128: Delay Asymmetry Detection/Measurement - Setup>>image:434430579187449857_5.20.png||alt="Figure 128" width="550"]]

17

18

At first, the GM and T-BC-1 were connected to the GNSS, and T-BC-2 used T-BC-1 as its primary time source. The GNSS was then disconnected from T-BC-1; T-BC-1 then locked onto the GM as its new time source, with the link between the GM and T-BC-1 being impaired by introducing an asymmetric delay of 500ns. During this process, the time error was measured on T-BC-2, and none of the measurements reflected the 500ns asymmetry, proving that all T-BC-1s could detect and eliminate the asymmetry.

19

The network passed the class accuracy level of 6 for all test runs, as per ITU-T G.8271, throughout the entire test run, including the switchover; the first three test runs, which used PTP profile ITU-T G.8275.2 throughout the entire chain all passed class accuracy level 6C. The IWF test run passed the class accuracy level of 6A.

20

21

These results are noteworthy for Time Synchronization. They prove that asymmetric delays can very well be handled, beating one of the big challenges of Time Synchronization.

22

23

24

(% class="table-bordered" %)

25

|=GM-A|=T-BC-1|=T-BC-2|=Impairment device|=Timing Test Equipment

26

|Calnex Paragon-neo PAM4|Microchip TimeProvider® 4500|Huawei ATN 910D-A||Calnex Paragon-neo PAM4

27

|Keysight Time Sync Analyzer|Ciena 5134|H3C S12500R-48Y8C||Keysight Time Sync Analyzer

28

29