Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [EFM] Wavelength allocation





[John Pickens]
> In the interest of taxonomy clarity we need a terminology that
> differentiates between actual PTP (via the physics of materials)
> and virtual PTP (via the physics of frequency spectra).
> Maybe MPTP and FPTP.

I'm having a terrible time trying to point out this difference to other
people at work. I propose this terminology:

PPTP, or PhyPTP - Physical point-to-point, which means a dedicated fiber
connecting each endpoint to the active distribution switch.

LPTP, or LogPTP - Logical point-to-point, which means a dedicated,
transparent, protocol-independent, logical channel over a shared medium.

> Also in the interest of taxonomy clarity we probably also need
> a terminology that differentiates between fixed and variable
> frequency assignments (since everything happens at a band of
> the electromagnetic spectrum), maybe fixed frequency (wavelength)
> vs frequency (wavelength) agile solutions.

This point was raised early today by a friend of mine. We must make a
difference between solutions that are:

- Fixed frequency only;
- Tunable by changing some hardware, by means of a plug-in replacement or
similar reconfiguration;
- Tunable by network management ('frequency agile', as you call).

The reason to have the intermediate category is to account for some
equipment, such as CPEs, where the cost for tunable lasers is not likely to
be viable for a long time. In the case of CPEs, frequency agility may be
easier to attain by changing some hardware or other hardware-related
configuration option. On the other hand, it may be economically feasible to
have tunable OLTs in a much shorter time frame.

> Whether frequency agility is undertaken as an element of TDMA EPON
> is a decision that needs to be made at some point, though is
> obviously influenced by the current relative cost of frequency
> (wavelength) agility.

I believe that at least some level of frequency agility must be part of the
standard. It is very important to make sure that first generation systems
use well-defined wavelength assignments, in order to make it possible to
have other wavelengths added as 'layers' over the same PON in the future.



Carlos Ribeiro
CTBC Telecom