The basic problem with 21CN is that it mixes two conflicting
agendas – IP and SLAs (Service Level Agreements). Notice that IMS (the
control layer) seems to have faded from the scene. This is part of the point I
make in http://frankston.com/?n=IPMarkeyIII
– as long as we entangle ourselves in the telecom mindset that is all
about supporting enumerated services rather than creating opportunity we’ll
keep tripping across ourselves.
Look at all the complexity in the summary. It’s a result of
trying to deal with all sorts of requirements in the plan. Things become far
simpler if we normalize to bits. That decoupling is the key enabler to the
hyper-growth that is lacking where telecom is able to assert control.
The current system is indeed collapsing – the question is
why we are trying so hard to prevent this collapse and thus deny us the benefit
if simple connectivity?
[mailto:firstname.lastname@example.org] On Behalf Of
Sent: Wednesday, August 26, 2009 14:09
Subject: [ NNSquad ] BT NGN runs into difficulty
NN squad members may interested in this UK Ofcom (the nearest we have to an
FCC) consultation on managing the consequences of BT slowing its investment
in NGN. From an engineering perspective some may say slowing equates to
a major failure, but either way it raises several questions for those
interested in securing an open bit commons as envisaged by Bob Frankston
>From a neutrality perspective I would welcome some reaction to this notion
of a control layer referenced in paragraph 1.10 which seems unworkable,
when all that should be needed is a transport layer with known emergent
properties. The latter is the key to maintaining open and neutral
networks going forward.
The reasons for the slowing / failure are many. The plan to emulate all
PSTN (POTS) features seamlessly over IP, so mass seamless migrations
could occur was certainly one challenge too many.
Either the telecomms engineers created a flawed design, or the disruptive
nature of an IP NGN was too big to embrace in full.
BT efforts to transfer all risk to its strategic box manufacturers (each
designed in isolation) while having no definitive picture of the end to
end service outcomes could be another factor, but I am guessing.
In my opinion the attempt to emulate fully the PSTN was a mistake,
and focus should be on delivering a transport layer where the properties permit
VOIP to work as it needs to using the existing TCP/IP mechanisms.
Comments and pearls of wisdom welcome, particularly on the need
for a Telco controlled control layer. I assume all incumbents will
struggle with these same issues, which begs the question as to how the
transition to a bit commons will occur.
one of the first to try Windows Live Mail.