RBS is often times run on on AMI D4 circuits. Also, RBS is generally
not run on inter-office trunks.
At least that was the case a few months ago when I worked on a DMS-500.
I have a feeling RBS is being used for inter-telco trunks (eg USwest to
GTE). And its killing v.90 in Seattle.
If a telco is running RBS on a trunk, any way to get them to remove it?
Also, any way to get the telco to tweak digital pads? Eg what magic
incantation do you say to the telco.
-Dan
RBS is often run on B8ZS/ESF too.
The reason it is not commonly run on inter-office trunks is because of
a remarkable new innovation called SS7. Perhaps you have heard of it.
Anyway, since the bits in robbed-bit signaling are "robbed" from the
superframe not from your DS0, it is irrelevant to your throughput.
---Rob
RBS is often times run on on AMI D4 circuits. Also, RBS is generally
not run on inter-office trunks.
At least that was the case a few months ago when I worked on a DMS-500.
rs@bifrost.seastrom.com ("Robert E. Seastrom") writes:
RBS is often run on B8ZS/ESF too.
The reason it is not commonly run on inter-office trunks is because of
a remarkable new innovation called SS7. Perhaps you have heard of it.Anyway, since the bits in robbed-bit signaling are "robbed" from the
superframe not from your DS0, it is irrelevant to your throughput.
Uh, no. If robbed bit signaling is used, every single DS0 gets
exactly one bit clobbered out of every six PCM samples (750us). Out
of 24 frames (an "Extended Super Frame"), those four stolen bits form
the familiar ABCD signaling pattern. (24 bit long pattern in the
framing bit signals the synchronization of the 24 frame sequence so
you can distinguish A from B, et cetera.)
(Perhaps you're thinking of FDL, which is carried purely over the
framing bit as a 4Kbps channel and doesn't affect any of the DS0s.)
Dan Hollis originally posted:
> > Is there any way to determine if inter-telco links are using
> > robbed bit
> > signaling? Eg from GTE to USwest in the Seattle area.
> >
> > I suspect RBS may be causing lots of v.90 problems in
> GTE country in
> > Seattle. But id like to know how to confirm this. And just
> > maybe, get it
> > fixed.
I doubt strongly that RBS would be the problem, since V.90 is designed
to work correctly with RBS channels. Two things to check:
- If they're using channel banks between you and the
customer's switch (like if they dropped you cheap T1s by
putting them line-side on your switch, rather than trunk-
side), then you lose. 24Kbps will be about the maximum.
- Some misconfigurations will merely cause lots of errors and
poor performance rather than abject failure. Check AMI
versus B8ZS (look for coding violations), and check for
clocking problems (extremely frequent slips noted by FDL).
someone here in our office was saying you can extract this out
of the modem after and during the call. Consult your modem manual
for specific directions, but I think they said it was at&v1
- jared
ATY11. USR modems only.
This gives a frequency response chart from which one can deduce whether
there are multiple A-D conversions in the line. I dont know if you can get
anything else from them though.
-Dan
Anyway, since the bits in robbed-bit signaling are "robbed" from the
superframe not from your DS0, it is irrelevant to your throughput.
This is actually not correct; my memory was at fault.
Robbed bit swipes the 8th bit of each timeslot during the 6th and 12th frame
for superframe; the 6th, 12th, 18th, and 24th frame for ESF.
Of course, you would not be reading this right now had
cskidmor@nwnexus.net not taken the liberty of re-adding NANOG to the
effort to keep the noise down on the list. 
---Rob