Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Networking The Internet

Irish Gov't Invests In Color-Coded Fiber Optics 129

c0mpliant writes "The Irish government has invested a further €5 million, after already having invested €5 million one year ago, in a new system of fiber optics which heralds an era of virtualization of fiber networks, using color coding to enable multiple fiber providers to serve businesses and homes, often on a single strand of fiber. The technology, which has already sparked interest from companies such as BT and IBM, is already in its first phase and boasts an impressive 2.5 terabytes capacity, double the capacity of the London phone system. The company behind the technology, Intune Technology, is comprised of a group of ex-UCD photonics researchers and has been around since 1999 and are based in Dublin. The project is set to be completed by 2020."
This discussion has been archived. No new comments can be posted.

Irish Gov't Invests In Color-Coded Fiber Optics

Comments Filter:
  • by 2phar ( 137027 ) on Friday July 16, 2010 @05:24AM (#32924140)
    Here's a Presentation from April [www.ria.ie] with some detail. There's more to this than just regular WDM.
  • Do they mean WDM? (Score:2, Informative)

    by 3.14159265 ( 644043 ) on Friday July 16, 2010 @05:31AM (#32924172)

    Intune Networks (...) has developed a technology that can enable a single strand of fibre to move from carrying one signal from one operator to carrying data from 80 telecoms and TV companies all at once."

    Do they mean they have "invented" http://en.wikipedia.org/wiki/Wavelength-division_multiplexing [wikipedia.org]? I'm puzzled.

  • Re:Terminology (Score:5, Informative)

    by pehrs ( 690959 ) on Friday July 16, 2010 @05:32AM (#32924178)

    What IS interesting is the idea of the fibre being shared by competing telcos. Has that been done before?

    Yes, it has. Selling wavelengths in dark fiber is very common, and companies frequently buy part of lines from eachother. Submarine cables are frequently owned by several companies.

    On a level closer to the customer there exists a (in Sweden) functional business model where a company owns the line to the customer and creates a market where different ISP's can provide services to the customer. OpenNet is one of more well known providers using this business model in Sweden.

  • Re:Terminology (Score:4, Informative)

    by fridaynightsmoke ( 1589903 ) on Friday July 16, 2010 @05:53AM (#32924256) Homepage
    Already happens in the UK. Multiple providers working off one cell tower is very common.
    Also common is 'virtual' cell networks, where the consumer-facing provider rents cell capacity from an established telco as you suggest. Examples of this are Tesco Mobile and Virgin Mobile, neither of which own or operate any infrastructure, instead piggybacking off other networks and offering their own pricing and service structure.
  • by cycoj ( 1010923 ) on Friday July 16, 2010 @07:18AM (#32924544)

    Wavelengths, we're pretty much always talking near-infrared. The most often used wavelength range is the so-called C-band (1530–1565 nm). This is mainly because this is where Erbium-doped fibre amplifiers work, which are necessary to create very long links without repeaters (also this is where the absorption minimum of fibres is). Less common is the L-band 1565-1625 nm. There's also the O and E band this are AFAIK mainly legacy bands which were used at the beginning of fibre optic communications. (Dense) Wavelength division multiplexing WDM/DWDM systems have channel spacings of 100 or 50 GHz bandwidth, i.e. ~1 or 0.5 nm at 1550 nm (look up ITU grid for more info). Today they usually carry 10 Gb/s data, however more and more 40 Gb/s and the next standard is 100 Gb/s (this actually uses multi-level phase and amplitude coding). About the pushing microwaves over optic lines, that's not really desired, absorption is too high, you also need diameters on the scale of a wavelength (depending on your index contrast) and finally you would not gain much, the bandwidth of your channels depends on your carrier frequency at microwave wavelengths the carrrier frequency is a couple of 100 GHz max. so that's the full bandwidth you get. compared to the C-band which contains about 70 channels at 100 GHz bandwidth each. So really no point.

  • Re:DWDM (Score:3, Informative)

    by Guy Harris ( 3803 ) <guy@alum.mit.edu> on Friday July 16, 2010 @07:54AM (#32924658)

    ...it's disappointing to find out that he is not an ex-UC Davis alumnus as the summary implies

    The summary speaks of "of a group of ex-UCD photonics researchers"; there's more than one UCD on the planet, and this one is probably University College Dublin [www.ucd.ie].

    The founders were from UCD, according to the "about Intune" page [intunenetworks.com]. ("Intune was founded in 1999 by two college graduates, John Dunne and Tom Farrell. They were performing research on tunable lasers and their network applications in University College Dublin, Ireland.")

    (I hope, for UCD's sake, that their Web designers aren't ex-UCD. Not only do they appear to think that "company" is spelled "copmany", the site is a Flash-infested mess that requires you to pop up several layers of menus, by clicking on little + signs, for each member of the management team whose biography you want to see.)

  • Re:DWDM (Score:3, Informative)

    by pz ( 113803 ) on Friday July 16, 2010 @08:17AM (#32924768) Journal

    The problem I see with this is the fact that certain wavelengths have certain interfering effects with other wavelengths. For example, 660-670nm radiation coupled with 720-740nm IR radiation causes some odd effects, which plants happen to utilize in photosynthesis, but I don't think we've ever tested such effects against the communication of data.

    What planet are you living on? Non-linear interaction between frequencies of EM communication has been studied for not just years, but decades. It's well understood. The subject is covered in any decent first-year Electrical Engineering course, and covered in much more detail in any decent course on Signals and Systems. Given linear media, supperposition applies, and there's no interaction. Given non-linear media, you get frequency mixing; with accurate knowledge of the non-linear characteristics, you have exact knowledge of the mixing. This isn't new Physics, it's Science that's so well understood it's become Engineering.

    The parent post is a clear example of the need for -1, Naive.

  • by pmullen ( 1858160 ) on Saturday July 17, 2010 @04:17AM (#32935594)
    In fact they are using a techique called lambda burst switching, where each switch on the ring has a colour associated with it and when sending a packet to that switch the transmit laser retunes to that wavelenght.

    It is very different than the usual OTN style DWDM transport and more like a giant ethernet switch, where every device gets 10Gbit/s uncontented transmit, that can send a packet directly to any other device, regardless of the number of devices in the ring. The number of nodes is only bound by how many lambdas you can fit in the C-band.

    They are basically statistically multiplexing in the time domain.

    The idea is cool but I'm not sure about the implementation.

Always draw your curves, then plot your reading.

Working...