PTP over DWDM transport

Dense Wavelength Division Multiplexing (DWDM) is a transmission technology that combines multiple light signals on a single fiber, resulting in higher fiber capacities, extended transmission distances, and higher data rates per channel.
It can be found in telecom networks, data centers & clouds, and enterprise networks and provides a scalable way to meet the growing data rate requirements.
This test case was designed to validate reliable PTP operation over DWDM optics for scalable, flexible, high-performance transport.

We tested three test runs with different vendor combinations for this test case.
For the first run, Ciena's QSFP-DD 400GE ZR+ tunable optics were configured to a wavelength of 1560.61 nanometers (nm) and a data rate of 400G. T-BC-1, the Juniper PTX10002-36QDD, was locked to PTP and SyncE from the Calnex Paragon-Neo PAM4 acting as an emulated T-GM, which also served as the measurement tool in this setup. T-BC-2, the Ciena 8140, was locked onto T-BC-1 over the DWDM interface on PTP and SyncE.
The Calnex Paragon-Neo PAM4 measured a peak-to-peak two-way time error of 3.782ns, with the average time error being 6.407ns and a constant time error (cTE) of 6.403ns for PTP on the output of the Ciena 8140.
For SyncE, a Maximum Time Interval Error (MTIE) of 0.25ns was measured, which is significantly below the ITU-T G.8262.1 "Cascaded eEECs Wander Generation Constant Temperature" mask.
     
For the second run, QSFP-DD  100GE ZR+ tunable optics were used. However, this time, it was configured to a wavelength of 1532.68 nm and a data rate of 100G. T-BC-1, the Ciena 5134, was locked to PTP and SyncE from the Keysight Time Sync Analyzer acting as an emulated T-GM, which also served as the measurement tool in this setup. T-BC-2, the Ciena 8140, was locked onto T-BC-1 over the DWDM interface on PTP and SyncE.
The Keysight TSA measured a peak-to-peak two-way time error of 6.766ns, with an average time error of 3.764ns and a cTE of 3.772ns on the output of the Ciena 8140 for PTP.
For SyncE, an MTIE of 3.4ns was measured, which is also significantly below the ITU-T G.8262.1 "Cascaded eEECs Wander Generation Constant Temperature" mask.
 .
    
QSFP-DD  800GE ZR+ tunable optics were used for the third and final test run; however, this time, they were configured to a wavelength of 1531.80nm and a data rate of 800G. T-BC-1, the Juniper PTX10002-36QDD, was locked to PTP and SyncE from the Calnex Paragon-Neo PAM4 acting as an emulated T-GM, which also served as the measurement tool in this setup. T-BC-2, another PTX10002-36QDD, was locked to T-BC-1 over the DWDM interface on PTP and SyncE.
The Calnex Paragon-Neo PAM4 measured a peak-to-peak two-way time error of 5.75ns, with the average being 0.158ns and a cTE of 0.158ns on the output of the second Juniper PTX10002-36QDD (T-BC-2) for PTP.
For SyncE, an MTIE of 1.25ns was measured, which is once again significantly below the ITU-T G.8262.1 "Cascaded eEECs Wander Generation Constant Temperature" mask.
    
These results are well below the threshold for T-BCs Class C media converter pairs, as defined in ITU-T G.8273.2 Appendix V, table V.1, with a requirement of +/- 20ns cTE per media converter pair and 45ns for the max absolute time error.