RE: [openss7] Re: [Megaco] Why separate Signaling Gateway from MGC?

["Dan Dunning" <dan.dunning@clarent.com>]

Here's one reason for you...

o Seperating the MGC from the SG allows the MGC to only support IP traffic -
it uses IP to control the Media Gateways (via MGCP typically) and IP to
control the SG (via SIGTRAN, TALI, or some proprietary IP based protocol).
There is no need to burden the MGC with SS7 cards, different physical
interfaces, etc., SS7 redundancy requirements, etc. Now I can select a MGC
platform without having to worry about which SS7 cards it supports, doing
compatability testing and prototyping, etc. Instead I leave much of the SS7
intricies to the SG which is supposedly implemented by a vendor with a lot
of experience, etc. in the SS7 and traditional TDM world. In the end, I am
able to get my MGC to market much sooner. BTW - I am referring to the lower
layers of the SS7 stack (SCCP/MTP3 and below). ISUP, TCAP, etc. can be dealt
with at the MGC if desired...

dan

> -----Original Message-----
> From: Patrick Lam [mailto:allmailinglist@yahoo.com]
> Sent: Monday, December 10, 2001 2:45 PM
> To: Tom-PT Taylor; openss7@openss7.org
> Cc: megaco@ietf.org
> Subject: [openss7] Re: [Megaco] Why separate Signaling Gateway from MGC?
>
>
> Tom:
>
> Thanks to your reply.
>
> You are right.  I didn't think about the sigtran group ..... sorry.
>
> However, I still have questions about why SG cannot be combined
> with MGC in
> one box.
>
> To me, combining them into one box provides cost-performance advantage
> (less transportation of the signaling, less possibility of errors, and
> cheaper to manufacture).
>
> Am I missing something?
>
> Thanks again in advance,
>
> Patrick.
>
>
> ----- Original Message -----
> From: "Tom-PT Taylor" <taylor@nortelnetworks.com>
> To: "'Patrick Lam'" <allmailinglist@yahoo.com>; <openss7@openss7.org>
> Cc: <megaco@ietf.org>
> Sent: Monday, December 10, 2001 9:09 AM
> Subject: RE: [Megaco] Why separate Signaling Gateway from MGC?
>
>
> > The question should have gone to Sigtran rather than Megaco, but I'll
> answer
> > it.
> >
> > The SG doesn't actually look at the call signalling: it does a transport
> > conversion for the layers beneath it.  Thus MTP3 becomes MxUA/SCTP/IP,
> Q.921
> > becomes IUA/SCTP/IP, and so on.  The signalling interworking is all done
> in
> > the MGC.
> >
> > The other point is that these are functional roles, so vendors
> are free to
> > terminate the original transport directly on the box supporting the MGC
> > function where that makes sense.
> >
> > -----Original Message-----
> > From: Patrick Lam [mailto:allmailinglist@yahoo.com]
> > Sent: Monday, December 10, 2001 12:12 AM
> > To: openss7@openss7.org
> > Cc: megaco@ietf.org
> > Subject: [Megaco] Why separate Signaling Gateway from MGC?
> >
> >
> > Hi:
> >
> > I hope these will be the right groups to ask this question ..... :)
> >
> > I have done some research on SS7 Signaling Gateways and MGCs.
> I found an
> > interesting similarities among most vendors:  the SG and MGC are usually
> > separated into physically differen boxes.  While I understand that the
> > scalability of the MGC might be the reason that the SS7 has to be
> separated
> > from it, but I really want to know why the vendors can't combine the
> > functionalities of SG and MGC together into one box?
> >
> > Why not have the SG decode all the ISUP and do MGCP and H.323 all in one
> > box?  What's the advantage of having a SG to convert SS7 to Sigtran, and
> > then having MGC convert Sigtran to H.323 and do MGCP?
> >
> > Thanks very much in advance,
> >
> > Patrick.
> >
> > _______________________________________________
> > Megaco mailing list
> > Megaco@ietf.org
> > http://www1.ietf.org/mailman/listinfo/megaco
>
>
> _________________________________________________________
> Do You Yahoo!?
> Get your free @yahoo.com address at http://mail.yahoo.com
>