[comp.dcom.telecom] ESF Framing Bits

kay@uunet.uu.net (Kay Fleskes) (10/12/90)

In article <13087@accuvax.nwu.edu> BRUCE@ccavax.camb.com (Barton F.
Bruce) writes:

>>the framing bit also (under ESF) can carry a small amount of
>>network managment data.

Specifically, 4Kbps worth of data.

and chip@chinacat.unicom.com (Chip Rosenthal) writes:

>That "facilities data link" seems to be a frequently underutilized
>feature.  Part of the problem is that different carriers and
>organizations (and vendors) have defined different standards for what
>should appear in this channel.  Another problem is that the carriers
>standards usually define a way for them to monitor performance of the
>user equipment, but don't provide users' the ability to monitor the
>performance of their network.  That is, the carriers are happy to take
>information, but not give it, and thus many folks are unhappy about
>that missing capability as well.

"Intelligent Diagnostic CSUs" are commonly used to make this
information available to the customer.  Different vendors support the
customer to different levels, but all venders are required to support
the minimum set of functionality identified in the T1.403 and 54016
Specs if they want to have the blessings of these people.
Additionally, the vendors may implement a superset of functionality,
usually to make themselves more attractive to their customers.

In our own case, we give the customer the ability to generate the same
requests that are being issued towards our equipment (which doesn't do
much good unless it is used in a point-to-point application, or in a
private network), AND, the ability to monitor the data that the
carrier can request from us.  Therefore, the customer has the ability
to monitor the service that is being provided to them by their
carrier.

>The big thing ESF gets you which is used universally is a six-bit CRC
>to provide a check on each 24-frame superframe.

The CRC error count is a much more accurate indicator than BPVs,
because it can detect logical errors in the data stream.  Both the
54016 and T1.403 specs require a measure of the level of these errors.
54016 requires a running count to be kept for retrieval by a data link
message, and T1.403 requires that within the Performance Response
message that is sent on the data link once per second, there is an
indicator of the error count for the previous four seconds.


Kay Fleskes     Design Engineer    ADC Kentrox   (...kentrox!kay)