Grumble flick
Only you guys can turn a story like this into the amount of porn and how fast you can squirt down a tube...
Alcatel-Lucent has teamed up with BT to test an "alien superchannel" across an existing fibre pipe, with the resulting record-breaking 1.4 Tbps they achieved able to transmit a five-month-long grumble flick in just one second. The pair achieved 1.4 terabits per second using a BT's fibre-optic pipe between the BT Tower and BT's …
It will benefit all as a lot of your traffic and everybody else's traffic will transit through london. Not so true as it used to be with peering points in Manchester and Leeds now and probably other places but still a lot of traffic will at some point pass through connectivity in London.
This has nothing to do with "making your broadband faster". This is about increasing the speed of pre-existing links by upgrading the bits at the end, so they get network upgrades without having to actually upgrade (much of) the physical network.
The point is to ease congestion on POPs/exchanges that have reached capacity and yet have no commercial incentive to upgrade. Its not to make your ADSL go fast in a little village somewhere.
Plus, if you ever did want your little village to have fast internet, then the backbone within the UK would have to be much much faster. Hence working on things like this, rather than trying to lay 50x as much fibre on trunk links.
Boom, headshot.
What's the physical limit on a single twisted pair? Somewhere in the region of 150mb/sec in ideal conditions, I'd assume, given than 100mb aggregate seems fairly relaible at the moment given adequate filtering.
I'm guessing we won't see speeds above 120mb down/30mb up until FTTP becomes defacto standard - which would be another massive infrastructure push, and require rather more work inside the home to terminate the fibre and convert it to a 'friendlier' format (CAT5/6) for home routing.
Still, at least it's more bandwidth for the backbone, eh?
I think purists can argue that the term 'broadband' is exactly right for this deployment, based upon the words intended original meaning (wiki quote:)
"The term broadband refers to the wide bandwidth characteristics of a transmission medium and its ability to transport multiple signals and traffic types simultaneously."
It's entirely true that "Joe Public" equates the term to access technology to the premise. But we know better, eh?
this actually means is that the bigwigs in charge of BT have figured out a way of shoving 5 times as much data down a fibre optic cable.
This research has cost BT umpteen millions... which is still cheaper than digging up the cables and laying 5 next to each other......therefore , fat bonuses all round to the senior managers, P45s all round for the research team as they're not needed any more and .1 of a second downloading before I hit my allowance for the day.
No, what it means is that some of the boffins in Acatel.Lucent's labs have developed a way to send three times as much data through a single transponder on an 1830PSS (up from the previous 400Gb/sec of current transponders) and BT have provided a first-in reference installation.
The important thing about this demonstration is not that it can be done, but that it has been done in a production environment as an upgrade to an existing product, rather than being an entirely new product line or a lab demo.
Disclaimer: Yes, I work for one of the parties involved in this demonstration and work with the 1830PSS product.
Not really a big deal as you do not lay a cable with single fibres, On a backbone cables can have over 200 fibres. It is common to put a cable with much much more fibre than you need. Laying a fatter cables with extra fibres adds very little cost. The big cost is putting in the ducts (tubes in the ground) If there is space in existing ducts, laying a new cable is 1 ot 2 K pounds - two or three man crew one day per Km regardless of the number of fibres. Splicing cables together takes time. But you can do that as you fill the cable. Fibre is basically free, as it is an asset share by many. The cost is local loop to the house.
>as you do not lay a cable with single fibres
That is today, there is much single core fibre infrastructure that dates from the 80's ...
Also you are forgetting that a lease-line provider will charge for the number of dedicated circuits/cables. Hence this potentially means that a client of BT can 'upgrade' their existing fibre between their Docklands and Ipswich datacentres say with little or no increase in leased line costs. Which may be an important consideration for some companies in the City where they have run out of duct space...
The art will be pricing this so that the cost of upgrading the ends is appreciably lower than the cost of deploying more fibre either directly or by purchasing more circuits from BT say...
...we're talking three times the throughput at the termination point (the customer), without any expensive and long-term road and pavement works to replace cabling and roadside boxes?
My only niggle is wondering how long it'll take the relevant parties to roll this out into the real world.
Still, when all's said and done, it's nothing to be sniffed at, when you think about it.
I could do with the French peer my ISP is currently routing through sorting out their damned router! Proxad.net's routing to a site I stream to has been b0rked for at least two weeks now with massive packet loss in the evenings. What are those French doing I wonder. Hogging the bandwidth with Grand Freres(sp)?
But my ISP tells me it is out of their control. :(
Google maps shows the distance between BT Tower and Adastral Park is 88miles (150Km). So saying the cable went between these two points is like saying the London Marathon is from Blackheath to The Mall.
Was this cable used by anything else? It is normally used or is it a special research cable? Where does the cable go? Does it tour around Cambridgeshire first? It could visit Sheffield en route and still have some slack.
Telecoms topology rarely uses a straight line in the trunk layer - there are buildings and things in the way. An end-to-end route us usually made up of sections of fibre patched together. There can't possibly be enough traffic between London and Ipswich to warrant building a direct and exclusive cable route. It's likely that there are a number of concentric rings around London and the route to Ipswich is a spur off of one of those. Try looking at the distance from London to Birmingham and from there out to Colchester, and from there on to Ipswich.
There can't possibly be enough traffic between London and Ipswich to warrant building a direct and exclusive cable route.
Funnily enough, there are actually several direct trunks to BT's main research facility. Almost like they would push all their daily traffic up there to test stuff. Crazy.