RE: [Sigtran] ISDN BA layer 1 activation

["Miesch, Thomas" <[e-mail]>]

Hi Ken 
That's allright, I appreciated the time you take to answer my questions
I added my comments below yours. 

Best regards
Thomas

>-----Original Message-----
>From: Ken Morneault [mailto:[e-mail]]
>Sent: Friday, March 26, 2004 8:30 PM
>To: Miesch, Thomas; [e-mail]
>Subject: RE: [Sigtran] ISDN BA layer 1 activation
>
>
>
>Thomas,
>
>Sorry for the late reply.  Please see below.
>
>Ken
>
>At 10:55 AM 3/22/2004 +0100, Miesch, Thomas wrote:
>>Hello Ken
>>Thank you for your answer.
>>But there still remains a question or a point where I would like some
>>clarification.
>>In europe normally the ISDN BRI is deactivated and will be activated
on
>>a call by call basis.
>>For originating calls there are no problems, the activation can be
done
>>without problems.
>>But for terminating calls we must activate the layer 1, before we can
>>send a message to the subscriber
>
>
>do you have originating and terminating backwards in your description?
I hope not, the originating call will be initiated by a subscriber
connected to the MG. Then the layer 1 activation is initiated by the
subscriber and layer 2 frames are sent only if the layer 1 is already
activated.
The terminating call is a call received by a subscriber connected to the
MG. So the layer 1 is deactivated and must activated first before any
Layer 2 or any DL-UNIT-DATA can be sent.

>
>
>>How can this be done?
>>One way would be we listen to the layer 3 setup message and if we
>>receive it we activate layer 1.
>>The setup will be sent in a DL-UNIT-DATA. Is this the way to go or are
>>there other recommend solutions?
>
>Sorry, I've never done a BRI implementation.  I would assume
>that the SG would bring layer 1 up when it received a DL-UNIT-DATA
>and the link was down.  Why does it have to look to see if
>it is a SETUP message?

Your are right it doesn't have to check for the Setup message. Even if
in a DL-UNIT-DATA are some other informations. The layer 1 must be
activated first before it can be transferred.
>
>>The activation can take several seconds (up to 7s), so the timeout for
>>an response to the setup message
>>must be large enough in the SG?
>
>The Q.931 stack has timer that deals with response to SETUP.
>
>
T303 is only 4s, for Layer 1 cold starts that is not enough time.
Because cold starts need more time. So the setup message will be sent to
the SG twice or even the setup of the call is canceled due to time out
of the layer 3 timer.

That is the reason that in the V5 standard there is also a Timer for
that. In the ETSI V5 ETS300 324-1 standard, for the value it is
referenced to the ETS300 012. That standard is almost the same as ITU-T
I.430. There is written that the activation time can even be in some
rare cases up to 15 or even 30s.

So the question stays how will be an activation handled on a call by
call basis which is interopable between different vendors of MG+SG and
MGC.
Because the Layer 3 timer is not long enough. 
I still believe the most compatible solution would be that the
activation is initated before the Layer 3 starts the call setup.

Can somebody comment which already has implemented an BRI MG. 


>>Best Regards
>>Thomas
>>
>>>-----Original Message-----
>>>From: Ken Morneault [mailto:[e-mail]]
>>>Sent: Friday, March 19, 2004 3:50 PM
>>>To: Miesch, Thomas; [e-mail]
>>>Subject: Re: [Sigtran] ISDN BA layer 1 activation
>>>
>>>
>>>
>>>Thomas,
>>>
>>>The MG is responsible for Layer 1
>>>activation.  It is out of scope for
>>>IUA.
>>>
>>>Regards,
>>>
>>>Ken
>>>
>>>
>>>At 03:39 PM 3/19/2004 +0100, Miesch, Thomas wrote:
>>>
>>>>
>>>>
>>>>Hello
>>>>
>>>>After reading of the megaco standards and the IUA RFC.
>>>>
>>>>The question remains how the ISDN BA layer 1 activation is done.
>>>>
>>>>I didn't find any informations about it.
>>>>
>>>>
>>>>
>>>>As I understand there was a proposed IL1A draft
(draft-bouwen-sigtran-
>>>il1a-01.txt) by Mr. Bouwen and Mr. Morneault.
>>>>
>>>>But this paper has expired and is no longer valid.
>>>>
>>>>Somewhere I read that that paper is not in the focus of the Sigtran
>>>working group
>>>>
>>>>and will be handled only as an extension for the IUA standard.
>>>>
>>>>But in the new draft of the IUA standard
>>(draft-ietf-sigtran-rfc3057bis-
>>>00.txt)
>>>>
>>>>I didn't saw any information about that issue.
>>>>
>>>>
>>>>
>>>>So the questions stay:
>>>>
>>>>How should the layer1 activation be handled?
>>>>
>>>>Who is responsible for the activation of the ISDN BA Layer 1, the
MGC
>>or
>>>the MG?
>>>>
>>>>If the responsibility is by the MGC, how the Layer 1 activation is
>>done?
>>>>
>>>>In which document the activation sequence is described?
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>Best Regards
>>>>
>>>>Thomas Miesch
>>>>
>>>>
>>>
>>
>>
>>_______________________________________________
>>Sigtran mailing list
>>[e-mail]
>>https://www1.ietf.org/mailman/listinfo/sigtran
>


_______________________________________________
Sigtran mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/sigtran