| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
David, RTT calculation is well defined in EPON and I do not see any need to modify that. I did not suggest any such changes. Marek From: David Gamliel [mailto:gamliel.david@xxxxxxxxx] Hello Marek, Thanks for the explanation. So as a matter of fact, do you suggest that the 802.3bf time stamping point will be used also for the RTT(i) calculation by the OLT in any EPON system?
On Mon, Apr 2, 2012 at 4:51 PM, Marek Hajduczenia <marek.hajduczenia@xxxxxx> wrote: David, I am confused about the context of this email. 802.3bf does not mandate any specific implementation of any IEEE 1588v2 profile, so you are free to choose 802.1AS or any other more generic IEEE 1588v2 profile you might personally prefer. Note that by itself, 802.1AS needs 802.3bf for precise timing indication. This was one of the reasons for 802.3bf project in the first place. So if you decide to use 802.1AS, you need to use 802.3bf as well. If you opt for other IEEE 1588v2 profile, you may add 802.3bf for improved precision. Marek From: David Gamliel [mailto:gamliel.david@xxxxxxxxx] Hello, In EPON, Time Of Day (ToD) recovery in the ONU can be done by the method defined in IEEE 802.1AS Clause 13: "Media Dependent Layer Specifications for EPON". The method standardize there seems simpler to implement and more accurate potentially then the deployment of IEEE1588 in EPoC. In addition, it does not require a servo based control system for ToD recovery, as required in 1588 systems. Why shoudn't this method (i.e. the one in 802.1AS clause 13 ) be deployed also for EPoC? Thanks, and BR, David |