IEEE 1952 PNT Resilience Levels: Resist
Five Minute Facts About Packet Timing
I’d like to tell you something about the next IEEE P1952 resilience level in our series.? If you missed it, I have been posting about the proposed resilience levels from the IEEE P1952 project.? Today’s topic is resilience level 3: Resist.
?For PNT equipment Resist means to continue working, with acceptable performance for some period time.? For Navigation that might mean dead reckoning assisted by accelerometers.? For timing system that means holdover with an oscillator.? Clocks in holdover get less accurate over time since the holdover oscillator is imperfect and drifts off frequency over time.? The amount time you can holdover with acceptable time error depends on the stability of the oscillator and the strictness of you applications timing requirements.? I described how holdover works and how to select a holdover oscillator in a previous blog post.
?If you have a network that supports Synchronous Ethernet (SyncE) then devices on the network can improve their holdover time by locking their local oscillator to the frequency of the SyncE source, which usually has a better local oscillator.? This is shown in Figure 1. ?As an aside, PTP time transfer accuracy generally improves when the PTP timeReceiver and GM are frequency locked at the physical layer.?
领英推荐
One problem with the network in Figure 1, is that if the reason the timeReceiver is in holdover is a problem with the GM, then SyncE might also be unavailable.? Figure 2 shows a PTP network where the source of SyncE is different than the PTP GM.? If the GM fails in this case, then the timeReceiver can still locks its local oscillator to the source of SyncE.
If you have any questions about network timing, don’t hesitate to send me an email at?[email protected], or visit our website at?www.meinbergglobal.com.
?
I suspect the need for an independent PRC is pretty important with frequency hopping communication networks, such as used to keep military comms secure from enemy capture and decryption. If the clock frequency at either the sending or receiving device drifted then frequency changes would be lost, or wouldn't be known, for a brief period. Thus some information would be dropped, never to be known to the receiving unit. Correct, Professor Doug?