T.R | Title | User | Personal Name | Date | Lines |
---|
955.1 | dark fibre in U.K | LARVAE::WILKINS_N | | Tue May 18 1993 04:01 | 16 |
|
The PTTs in the U.K will not supply dark fibre, their licence does not
permit it and they have no reason to want to change that. Their
charges are on bandwidth. They must terminate lines with some form of
electrical/optical equipment.
We have been working with the PTTs for a number of years to encourage
them to supply FDDI leased lines using our equipment on the ends. Hence
the line we rent from DECpark to Acre road. A number of other companies
also rent a service mainly for MDF usage.
BT should be launching a service shortly. Not an ideal situation, but
the other way is to keep going to court to try and get the rules
changed.
Neil
|
955.2 | Time to reopen the debate ? | BROUGH::DAVIES | Not Also, but ONLY | Tue May 18 1993 04:23 | 24 |
| The radio prog did indeed mention that fact. However the debate that led to
the current situation took place in the early/mid 1980's. Since the invention
at Southampton University of the iridium doped optical amplifier, the situation
has changed radically.
The US cable companies operating franchises in the UK do seem prepared to
extend the cable from the street into the consumers homes when quote "they
can handle the bandwidth".
If Digital was able to setup Dark Fibre links to employees homes then the
acres of car parking at most digital offices could be done away with. You could
send over the link your network link,DTN, video Phone, DVN pictures etc. Real
Homeworking at last.
You are correct about the licensing issue. BT is a problem but Mercury is
pressing for a change. Lord Young (chairman of C&W, Mercury parent co) and
a former Government Minister spoke stongly that the debate should be re-opened.
My original question was really relating to the development of the equipment
to collect/distribute the data sources at each end of the Dark Fibre cable.
Stephen Davies
|
955.3 | inefficient use of technology, resource | DELNI::GOLDSTEIN | Bad cop | Tue May 18 1993 09:57 | 32 |
| Dark Fibre is to telecommuting as bullet trains are to pleasure drives.
While fiber optic strands are growing cheaper, they're still more per
foot than copper. What makes optics so favorable is its high
bandwidth. PTTs take advantage of that by using mulitplexors. The
current generation of SDH fiber optic gear will put 2.44 Gbps onto a
single strand. The muxes, of course, aren't terribly cheap because
electronic gear at gigabit rates isn't simple.
Dark fiber is basically a hack. It takes this nearly
infinitely-expandable resource and uses it all up for one thing. You
couldn't afford it to your house; "fiber to the home" projects
are mostly talking about 155 Mbps short spurs off of neighborhood
multiplexors. Dark fiber is what you need when you run a protocol like
FDDI that was designed for intra-campus LOCAL AREA use and want to
extend it across the PTT's franchise boundary.
We are facing a resource crunch on some cross sections of the Digital
Lightwave Network here in New England because we've allowed some of our
spare capacity to be sold as "dark fiber" to groups that are running 10
Mbps Ethernet on it. If it were not sold "dark", we could run dozens
of Ethernets' worth of traffic down it, via serializing bridges. But
LAN folks don't understand or want to deal with that technology.
US telcos have provided some dark fiber and are prohibited from
arbitrarily taking it away. That's reasonable; if the customer wants
to pay full cost let them have it. But in the long haul it's not very
economical except for short spans (like some of the Reading runs) where
wasting some strands is cheaper than buying muxes. Once you get into
optical amp distances (and you don't need them unless you're going
quite far, many km), it's a cryin' shame to waste the whole bandwidth
on one user, unless that user is really big.
|
955.4 | A good wind up! | ROCKS::CAMP | | Mon May 24 1993 09:15 | 28 |
| I also listened to the program regarding the use of fibre and the
techno revolution when used with the "near zero noise optical
amplifier", (I thought they said it was Urbium doped...), and it was
interesting. However one item was that they mentioned was the limitation of
the electrical intereface, but when on so say that that was now gone
with the high bandwitdh optics. The Catch 22 is that you need the
electronics before you can get it to the fibre!!! So how can you get
Terabits/sec down a fibre at one wavelength, if you can't get
Terabits/sec at the electrical interface. Perhaps they viewed the use
of multiple wavelengths each carring Gigabits/sec.
The point they were making is that with dark fibre the user is free to
push what he likes up the fibre, as fast as it can go (safely that is)
without having to cowtow to any standards, ie the user was free to do
what he wants, (yes I agree it could be wastfull, but then again you don't
use the full band width of your own domestic phone wire in any event).
The fact that TELCO's don't want to see access to dark fibre
because they currently get a good rake off from the bit rate and
equipment charges for the electrical connection doesn't help either
so they are unlikely to want to sell dark fibre. Maybe they will be
forced into it, even if its only for Metropolitan networks.
The best bit about it was that the reader was making a living talking
and writing about these fibre communications, which I thought was quite a
smart move, and winding up a lot of people in the process.
|
955.5 | | KONING::KONING | Paul Koning, A-13683 | Mon May 24 1993 12:20 | 24 |
| That's "Erbium".
You've got to remember that there is a big difference between:
1. The theoretical limit on fiber capacity (from first principles of physics
and information theory).
2. The current practical (state of technology) limit on fiber transmission
system capacity.
3. The current practical limit of channel bandwidth a customer can reasonably
use and/or afford.
There's at least a factor 100 between 1 and 2. There may well be another factor
of 10 to 1000 between 2 and 3. That one depends on what you do: supercomputer
users would answer differently than more "mainstream" commercial users.
Customers of US telecomm providers would answer differently than telecom
users elsewhere (competition vs. government monopolies); and so on.
Bandwidth is not free. Interfaces to infinite-capacity fibers are expensive.
You can't justify them on a "wouldn't it be neat to have them" basis.
That's where the gap between 2 and 3 comes from.
paul
|
955.6 | Telco's and Dark Fibre. | ROCKS::CAMP | | Tue May 25 1993 09:48 | 33 |
| Point taken regarding the bandwidth, and Erbium, (misplaced my periodic
table you see...)
I think that the TELCO's will be forced to supply dark fibre in the
long run, although they may not like to do it, even if its just
because they can't supply or maanage the associated Fibre/Electrical
interface. For example say a company wants to use FDDI in
London, the cost of managing each WC at the interface between the
company and BT "Fibre" would be prohibitive, irrespective of the
maintenance of the equipment. Now say we have 50 such companies each
with 10 nodes in differnt locations, thats 500 WC's at a minimum that
BT would have to support.
From past experience BT do not supply superfast fixes for hard failures
and less so for intermittant problems, (they are good for their own
trunk services though). I would also doubt their capabilities for
trouble shooting new brought in technology and also the compatability
of multivendor issues it could create. So I would see BT not
having much choice but to supply dark Fibre in the long run.
If a certain "American" bank in London can force BT to lay Dark Fibre
for its Ethernet Bridges, I see no reason why this will not be repeated
for other companies, newer technology and faster speeds.
An interesting space.
Mike.
I agree that its still matched to need and thats going to drive Dark
Fibre more than anything else.
|
955.7 | bt/dark fibre | LARVAE::WILKINS_N | | Wed May 26 1993 04:16 | 18 |
|
Mike,
Management is a "thorny" issue for BT, If I remember right BTs business
plan had no more than about 25 customers a year for a maximum of 3 year
service offering. AT the price they want to charge I doubt they will
get that.
They do not intend to manage every concentrator only one per service
and some form of SNMP to SMT gateway. Come in via slip and set traps in
the cons round the ring to alarm back at a central location.
You are right they don't have the skills to manage/service the kit, but
we do sub-contracted back.
They will supply "dark fibre" on a special project basis, but I guess
they just charge so much than nobody can afford it.
Neil
|