I believe it's IP/PPP/ATM/DSL to the home, and IP/Tag/ATM/Sonet in the
backbone.
That is about as succinctly put as can be, but I didn't get the impression
that all apps would run over one stack. There seems to be room for an ATM
API in the announcement, but no explicit mention.
The most radical change for an established telco like Sprint is that Sprint
has seen the light from the Great Crossover ahead and decided it is best and
cheapest to run voice calls over packet or cell than over separate switched
networks. It's getting to be a no-brainer decision, unless the guys in the
domed building in WDC eliminate switched network access charges to the
incumbents, which is apparently unlikely. So all voice moves to packet/cell
ASAP to avoid switched network call termination charges as much as possible.
Really a forced migration in terms of business costs, since the termination
charges in the one to three cent/minute range are enormous compared to the
cost of a packet voice call.
Are there any expected dates for deployment?
Given the timeframe to get multi-megabit copper or DSL from the incumbents
and the fact that within that timeframe qos-IP/DWDM/fiber equipment will
cost about 10% of IP/ATM/SONET/DWDM/fiber, I would say "never" is most
likely. Of course, they can deliver this to "big corporate customers" on
bypass right now, so perhaps the answer is "now". But "never" to the small
business and residential market.
The most radical change for an established telco like Sprint is that Sprint
has seen the light from the Great Crossover ahead and decided it is best and
cheapest to run voice calls over packet or cell than over separate switched
networks. It's getting to be a no-brainer decision, unless the guys in the
domed building in WDC eliminate switched network access charges to the
incumbents, which is apparently unlikely. So all voice moves to packet/cell
ASAP to avoid switched network call termination charges as much as possible.
Really a forced migration in terms of business costs, since the termination
charges in the one to three cent/minute range are enormous compared to the
cost of a packet voice call.
Right now most of the move to voice over packet/cell networks is
because people want to avoid termination and USF charges. The pace will
accelerate even more now that port density and port cost is approaching
that of TDM. If you wanted to buy a VOIP gateway 6 months ago you could
get around 24 - 96 ports with a port cost of around 2K. New hardware is
now becoming available that give you port densities greater then 2K with a
port cost of under $300 with SS7 support. In TDM networks you see port
densities as large as 100K with a port cost around $100. I don't think it
will be long until we see the same port densities with a port cost that is
under $100.