Discussion:
[Asterisk-Users] T1 E&M vs PRI question
Matt Beebe
2005-01-24 21:47:16 UTC
Permalink
Ok,

I'm about to take the plunge, and am trying to decide between Channelized T1 E&M and PRI. I'm getting an "Integrated T1" which will have data and voice capability, all plugged directly into my digium single T1 card. In either case the data piece looks pretty straighforward, just setup the channel properly, hand it off to the linux hdlc layer, and route away.... the voice side seems a little more complex -- I'm looking for clarification and/or advice:

It seems to me that the major differences between the two different voice delivery mechanisms (other than cost) is caller id functionality and call setup delay. With the PRI, I'll have practically instant call setup and the ability to pass CNAM (caller name) and CID (caller ID) information in BOTH directions. The PRI will give me the ability to have additional directory numbers (typically called DIDs) assigned against my voice trunks and will provide the full ANI (automatic number identification) and DNIS (dialed number identificaton service) over the PRI signalling trunk. Each voice channel will also be 64k clear channel, so I could (theoretically) provide 56k dial-in modem service from the same box (anyone actually doing this?? seems like a neat application for the dsp software guys) I also lose one 64k channel to signalling.

Sounds like the way to go, but basically the PRI ends up being $100/month more expensive than the Channelized T1 E&M.

The T1 E&M approach will still give me CID (but not CNAM???) over the in-band call setup mechanism (ie: quick DTMF tones during the wink). Each voice channel will actually be 56k because it uses RBS (robbed bit signalling -- not sure what its using this for, as the call setup is delivered via wink???). As a result, this approach would also keep me from implementing a 56k dial-in modem service, but I could still use an "ordinary" modem or fax dsp to provide 33.6k dial-in. This setup can support DID, but its appended (or prepended, depending on the provider) to the DTMF call setup (which extends the time for calls to actually connect). Not sure if CID or CNAM can be provided for outgoing calls (I think some providers can enable me to be able to wink to them the number to pass as caller id??)

I believe in either case, the normal call features (3-way, forwarding, etc) can be provisioned.

Do I have it about right?? Is it pretty normal for providers to charge a premium for the PRI? Any thoughts/clarifications to my above assumptions?? Are there other pros/cons of each setup?

Thanks in advance!

-Matt
Keith Burns
2005-01-24 23:49:57 UTC
Permalink
Depending on the switch they are using, there are a limited number of
D-channels (or D-channel licenses).

CAS signaling needs RBS (it's the winking in this case).


-----Original Message-----
From: asterisk-users-***@lists.digium.com
[mailto:asterisk-users-***@lists.digium.com] On Behalf Of Matt Beebe
Sent: Monday, January 24, 2005 2:47 PM
To: asterisk-***@lists.digium.com
Subject: [Asterisk-Users] T1 E&M vs PRI question

Ok,

I'm about to take the plunge, and am trying to decide between
Channelized T1 E&M and PRI. I'm getting an "Integrated T1" which will
have data and voice capability, all plugged directly into my digium
single T1 card. In either case the data piece looks pretty
straighforward, just setup the channel properly, hand it off to the
linux hdlc layer, and route away.... the voice side seems a little more
complex -- I'm looking for clarification and/or advice:

It seems to me that the major differences between the two different
voice delivery mechanisms (other than cost) is caller id functionality
and call setup delay. With the PRI, I'll have practically instant call
setup and the ability to pass CNAM (caller name) and CID (caller ID)
information in BOTH directions. The PRI will give me the ability to
have additional directory numbers (typically called DIDs) assigned
against my voice trunks and will provide the full ANI (automatic number
identification) and DNIS (dialed number identificaton service) over the
PRI signalling trunk. Each voice channel will also be 64k clear
channel, so I could (theoretically) provide 56k dial-in modem service
from the same box (anyone actually doing this?? seems like a neat
application for the dsp software guys) I also lose one 64k channel to
signalling.

Sounds like the way to go, but basically the PRI ends up being
$100/month more expensive than the Channelized T1 E&M.

The T1 E&M approach will still give me CID (but not CNAM???) over the
in-band call setup mechanism (ie: quick DTMF tones during the wink).
Each voice channel will actually be 56k because it uses RBS (robbed bit
signalling -- not sure what its using this for, as the call setup is
delivered via wink???). As a result, this approach would also keep me
from implementing a 56k dial-in modem service, but I could still use an
"ordinary" modem or fax dsp to provide 33.6k dial-in. This setup can
support DID, but its appended (or prepended, depending on the provider)
to the DTMF call setup (which extends the time for calls to actually
connect). Not sure if CID or CNAM can be provided for outgoing calls (I
think some providers can enable me to be able to wink to them the number
to pass as caller id??)

I believe in either case, the normal call features (3-way, forwarding,
etc) can be provisioned.

Do I have it about right?? Is it pretty normal for providers to charge
a premium for the PRI? Any thoughts/clarifications to my above
assumptions?? Are there other pros/cons of each setup?

Thanks in advance!

-Matt
David Boyd
2005-01-25 02:11:50 UTC
Permalink
Responses embedded below!
Post by Keith Burns
Depending on the switch they are using, there are a limited number of
D-channels (or D-channel licenses).
CAS signaling needs RBS (it’s the winking in this case).
-----Original Message-----
Sent: Monday, January 24, 2005 2:47 PM
Subject: [Asterisk-Users] T1 E&M vs PRI question
Ok,
I'm about to take the plunge, and am trying to decide between
Channelized T1 E&M and PRI. I'm getting an "Integrated T1" which will
have data and voice capability, all plugged directly into my digium
single T1 card. In either case the data piece looks pretty
straighforward, just setup the channel properly, hand it off to the
linux hdlc layer, and route away.... the voice side seems a little
PLease no Flame, just a correction if required.

There seemed to be issue using Data/Voice on the digium cards, but I
believe it is a setup issue not a technical limitation on the card
itself.
Post by Keith Burns
It seems to me that the major differences between the two different
voice delivery mechanisms (other than cost) is caller id functionality
and call setup delay. With the PRI, I'll have practically instant
call setup and the ability to pass CNAM (caller name) and CID (caller
ID) information in BOTH directions. The PRI will give me the ability
to have additional directory numbers (typically called DIDs) assigned
against my voice trunks and will provide the full ANI (automatic
number identification) and DNIS (dialed number identificaton service)
over the PRI signalling trunk. Each voice channel will also be 64k
clear channel, so I could (theoretically) provide 56k dial-in modem
service from the same box (anyone actually doing this?? seems like a
neat application for the dsp software guys) I also lose one 64k
channel to signalling.
Actually DNIS can be provisioned over e&m trunking also, the separation
of digits is done with *'s or KP/ST. So the digiti dump would be
something like:
DTMF
OH ->

<- Wink

digit dump *703727131229*8004231212*->
<-wink
<-Answer

The breakdown of the digits is ani + Info digits then DNIS

The *'s would be replaced with KP/ST pulses if MF. KP start sequence, &
ST stop sequence.

Sorry for the crude drawing, and the disclaimer is its been 4 years
since I have looked at the digit sequence for an E&M t1 :)
Post by Keith Burns
Sounds like the way to go, but basically the PRI ends up
being $100/month more expensive than the Channelized T1 E&M.
The T1 E&M approach will still give me CID (but not CNAM???) over the
in-band call setup mechanism (ie: quick DTMF tones during the wink).
Each voice channel will actually be 56k because it uses RBS (robbed
bit signalling -- not sure what its using this for, as the call setup
is delivered via wink???). As a result, this approach would also keep
me from implementing a 56k dial-in modem service, but I could still
use an "ordinary" modem or fax dsp to provide 33.6k dial-in.
This setup can support DID, but its appended (or prepended, depending
on the provider) to the DTMF call setup (which extends the time for
calls to actually connect). Not sure if CID or CNAM can be provided
for outgoing calls (I think some providers can enable me to be able to
wink to them the number to pass as caller id??)
I don't know of a way for outbound or inbound CNAM to be provided on a
T1 unless you are using SS7 or some like control protocol.

The setup time is in milliseconds for PRI and potentially could be 1.2
seconds in E&M including wink times, and outpulse dump. This can be
decreased if the carrier can accept fast outpulse, and also be decreased
if you use MF with KP & ST pulses instead of DTMF.

Robbed bit allows for the current channel condition to be maintained in
the signalling stream. When a channel hangs up the onhook condition has
to be able to be passed to the other end of the t1 for disconnect. The
wink and digits dump at the start of the call only provides call setup
capability.
Post by Keith Burns
I believe in either case, the normal call features (3-way, forwarding,
etc) can be provisioned.
Additional features are usually handled within the switching/* system
once the call has been setup. There are some features that are available
via ISDN, however in my experiences most carriers don't/won't support
them.
Post by Keith Burns
Do I have it about right?? Is it pretty normal for providers to
charge a premium for the PRI? Any thoughts/clarifications to my above
assumptions?? Are there other pros/cons of each setup?
Yes it is normal for increased cost, however IMHO I would spend the
additional money (assuming one can afford it) for improved throughput
and performance.
Post by Keith Burns
Thanks in advance!
-Matt
______________________________________________________________________
_______________________________________________
Asterisk-Users mailing list
http://lists.digium.com/mailman/listinfo/asterisk-users
http://lists.digium.com/mailman/listinfo/asterisk-users
Keith Burns
2005-01-25 02:19:19 UTC
Permalink
Correct, CAS can supply DNIS but the call set up times are significantly
longer.
Post by Keith Burns
-----Original Message-----
Sent: Monday, January 24, 2005 7:12 PM
To: Asterisk Users Mailing List - Non-Commercial Discussion
Subject: RE: [Asterisk-Users] T1 E&M vs PRI question
Responses embedded below!
Post by Keith Burns
Depending on the switch they are using, there are a limited number
of
Post by Keith Burns
Post by Keith Burns
D-channels (or D-channel licenses).
CAS signaling needs RBS (it's the winking in this case).
-----Original Message-----
Beebe
Sent: Monday, January 24, 2005 2:47 PM
Subject: [Asterisk-Users] T1 E&M vs PRI question
Ok,
I'm about to take the plunge, and am trying to decide between
Channelized T1 E&M and PRI. I'm getting an "Integrated T1" which
will
Post by Keith Burns
Post by Keith Burns
have data and voice capability, all plugged directly into my digium
single T1 card. In either case the data piece looks pretty
straighforward, just setup the channel properly, hand it off to the
linux hdlc layer, and route away.... the voice side seems a little
PLease no Flame, just a correction if required.
There seemed to be issue using Data/Voice on the digium cards, but I
believe it is a setup issue not a technical limitation on the card
itself.
Post by Keith Burns
It seems to me that the major differences between the two different
voice delivery mechanisms (other than cost) is caller id
functionality
Post by Keith Burns
Post by Keith Burns
and call setup delay. With the PRI, I'll have practically instant
call setup and the ability to pass CNAM (caller name) and CID
(caller
Post by Keith Burns
Post by Keith Burns
ID) information in BOTH directions. The PRI will give me the
ability
Post by Keith Burns
Post by Keith Burns
to have additional directory numbers (typically called DIDs)
assigned
Post by Keith Burns
Post by Keith Burns
against my voice trunks and will provide the full ANI (automatic
number identification) and DNIS (dialed number identificaton
service)
Post by Keith Burns
Post by Keith Burns
over the PRI signalling trunk. Each voice channel will also be 64k
clear channel, so I could (theoretically) provide 56k dial-in modem
service from the same box (anyone actually doing this?? seems like a
neat application for the dsp software guys) I also lose one 64k
channel to signalling.
Actually DNIS can be provisioned over e&m trunking also, the
separation
Post by Keith Burns
of digits is done with *'s or KP/ST. So the digiti dump would be
DTMF
OH ->
<- Wink
digit dump *703727131229*8004231212*->
<-wink
<-Answer
The breakdown of the digits is ani + Info digits then DNIS
The *'s would be replaced with KP/ST pulses if MF. KP start sequence,
&
Post by Keith Burns
ST stop sequence.
Sorry for the crude drawing, and the disclaimer is its been 4 years
since I have looked at the digit sequence for an E&M t1 :)
Post by Keith Burns
Sounds like the way to go, but basically the PRI ends up
being $100/month more expensive than the Channelized T1 E&M.
The T1 E&M approach will still give me CID (but not CNAM???) over
the
Post by Keith Burns
Post by Keith Burns
in-band call setup mechanism (ie: quick DTMF tones during the wink).
Each voice channel will actually be 56k because it uses RBS (robbed
bit signalling -- not sure what its using this for, as the call
setup
Post by Keith Burns
Post by Keith Burns
is delivered via wink???). As a result, this approach would also
keep
Post by Keith Burns
Post by Keith Burns
me from implementing a 56k dial-in modem service, but I could still
use an "ordinary" modem or fax dsp to provide 33.6k dial-in.
This setup can support DID, but its appended (or prepended,
depending
Post by Keith Burns
Post by Keith Burns
on the provider) to the DTMF call setup (which extends the time for
calls to actually connect). Not sure if CID or CNAM can be provided
for outgoing calls (I think some providers can enable me to be able
to
Post by Keith Burns
Post by Keith Burns
wink to them the number to pass as caller id??)
I don't know of a way for outbound or inbound CNAM to be provided on a
T1 unless you are using SS7 or some like control protocol.
The setup time is in milliseconds for PRI and potentially could be 1.2
seconds in E&M including wink times, and outpulse dump. This can be
decreased if the carrier can accept fast outpulse, and also be
decreased
Post by Keith Burns
if you use MF with KP & ST pulses instead of DTMF.
Robbed bit allows for the current channel condition to be maintained
in
Post by Keith Burns
the signalling stream. When a channel hangs up the onhook condition
has
Post by Keith Burns
to be able to be passed to the other end of the t1 for disconnect.
The
Post by Keith Burns
wink and digits dump at the start of the call only provides call setup
capability.
Post by Keith Burns
I believe in either case, the normal call features (3-way,
forwarding,
Post by Keith Burns
Post by Keith Burns
etc) can be provisioned.
Additional features are usually handled within the switching/* system
once the call has been setup. There are some features that are
available
Post by Keith Burns
via ISDN, however in my experiences most carriers don't/won't support
them.
Post by Keith Burns
Do I have it about right?? Is it pretty normal for providers to
charge a premium for the PRI? Any thoughts/clarifications to my
above
Post by Keith Burns
Post by Keith Burns
assumptions?? Are there other pros/cons of each setup?
Yes it is normal for increased cost, however IMHO I would spend the
additional money (assuming one can afford it) for improved throughput
and performance.
Post by Keith Burns
Thanks in advance!
-Matt
____________________________________________________________________
__
Post by Keith Burns
_______________________________________________
Asterisk-Users mailing list
http://lists.digium.com/mailman/listinfo/asterisk-users
http://lists.digium.com/mailman/listinfo/asterisk-users
_______________________________________________
Asterisk-Users mailing list
http://lists.digium.com/mailman/listinfo/asterisk-users
http://lists.digium.com/mailman/listinfo/asterisk-users
Andrew Kohlsmith
2005-01-25 02:15:48 UTC
Permalink
Please don't post to the list using HTML.
Post by Matt Beebe
in-band call setup mechanism (ie: quick DTMF tones during the wink). Each
voice channel will actually be 56k because it uses RBS (robbed bit
signalling -- not sure what its using this for, as the call setup is
delivered via wink???). As a result, this approach would also keep me from
implementing a 56k dial-in modem service, but I could still use an
"ordinary" modem or fax dsp to provide 33.6k dial-in. This setup can
This is not true -- I helped run a small ISP for years on CAS T1 (robbed bit
signaling) lines and could achieve 53kbps without even trying. 56k modems
already detect the timeslots and can "route around" the 1/6 frame "damage"
that inband signaling takes.

-A.

Loading...