Cell Relay Archive

[James Carlson <carlson@ironbridgenetworks.com>]

"ronald h. davis" <ronaldd@lucent.com> writes:
> there is, of course, another option and that is to use aal-2 with
> silence suppression.

Which serves only to make the traffic more bursty, which is exactly
what ATM excels in not handling correctly.

> as you suggested, aal-1 for voice makes more sense than aal-5 for voice
> because aal-1 has less overhead.  in fact, voice over atm is more
> efficient than voice over ip even when using aal-5.

Uh, I suggested no such thing.

Voice-over-AAL-1 might be considered to have slightly less overhead
than voice-over-IP-over-AAL5 iff you ignore G.723.1 and friends and if
you ignore ATM signaling overhead.  This is probably a complete wash,
though.

ATM, though, maintains state through each switch for the VPI+VCI
mappings plus the PNNI routing.  This state grows linearly with the
number of connections through the switch for the VC mappings and
roughly logarithmically with the size of the network.

IP has state for routing only, so it does NOT need per-flow storage,
and employs more sophisticated summarization techniques on the routing
as well.  (PNNI is like OSPF, but there's no ATM equivalent of BGP.)

Things are much better still for IP if you rip out the useless ATM
layer and run IP over SONET, or, better still, IP directly on fiber.

> voice really is the killer app for atm because it carries voice better
> than ip does...the same doesn't hold true for data.

I can agree with the data part, especially for very high speed
(~1Tbps) networks where the typically small buffering in ATM switches
interacts very poorly with the bursty nature of network traffic and
the AAL-5 SAR.  Folks playing with data over "voice grade" switches
are finding this a hassle.

Whether voice carried in individual VCs (!) as suggested is a "killer"
application for ATM remains to be seen.  I (and many others) still say
that it doesn't scale well at all.

About the only solid argument I've seen in favor of ATM is that,
unlike IP, telcos know how to use it now.  It's integrated into TRKS
and TMN.  Of course, there are some good integrated CMIP/SNMP
management tools available today, and things are still improving, so I
doubt this meager advantage will hold long.

-- 
James Carlson, Consulting S/W Engineer  <carlson@ironbridgenetworks.com>
IronBridge Networks / 55 Hayden Avenue  71.246W    Vox:  +1 781 372 8132
Lexington MA  02421-7996 / USA          42.423N    Fax:  +1 781 372 8090
"PPP Design and Debugging" --- http://people.ne.mediaone.net/carlson/ppp