Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

RE: [EFM] EFM Requirements


-----Original Message-----
From: Matt Squire
To: Harry Hvostov
Cc: Lough, Andy; Stds-802-3-Efm (E-mail)
Sent: 8/20/01 6:09 PM
Subject: Re: [EFM] EFM Requirements

Harry Hvostov wrote:
> Matt,
> For the record, the actual stream transport for VoIP is done via
> application
> layer protocols. Needless to say, these protocols would rely on the
> mechanisms supported at L2/L3. For an example of mapping the MGCP
> descriptions into L2 QoS parameter
> sets please see PacketCable 1.0 Dynamic QoS specification at
> The spec is publicly available.

So what.  At the Ethernet level a packet has priority.  Someone has
classified this packet as important.  Therefore, the network should do
what it can to not discard the packet and to give it a relatively low
latency.  The packet can be carrying voice, video, or my mother's secret
lasagna recipe - I don't care.  I've been told its important - the
application is irrelevant.  

And there are oodles of ways of carrying voice over Ethernet, some of
which use IP and some of which don't.  RTP is one possibility.  There is
also MPLS, circuit emulation, and numerous proprietary protocols.  And
again, its irrelevant.  Transport it using RTP in ATM in MPLS in CEM in
TCP in UDP in IP in Etherent for all I care - set the Ethernet priority
and we're all happy campers at L2.   

The point is that the Ethernet layer has no additional work to do to
carry voice.  Therefore the requirements for transporting voice need not
be singled out - there are none.   

> Harry
> -----Original Message-----
> From: Matt Squire [mailto:mattsquire@xxxxxxx]
> Sent: Friday, August 17, 2001 7:07 PM
> To: Lough, Andy
> Cc: Stds-802-3-Efm (E-mail)
> Subject: Re: [EFM] EFM Requirements
> Everyone is exploring SIP as a signaling protocol.  So what.  History
> littered with failed explorations.
> Might it happen?  Certainly.  Should we be able to support it if it
> does?  Sure.  But SIP is the signaling protocol, it has nothing to do
> with the data flow except to set it up.  Our L2 needs to have
> and be able to transport low loss/low latency data.  Its Ethernet, it
> can.  There is nothing for us to do as far as SIP and Ethernet are
> concerned.
> - Matt
> "Lough, Andy" wrote:
> >
> > Bob,
> >
> > my understanding is that most major carriers in EU are exploring SIP
> IP as their packet voice protocol going forward. So I believe this
> is made, our focus should be to provide the best layer1/2 package to
> this.
> >
> > R's
> > Andy
> >