Slashdot Mirror
Scientists Overcome One of the Biggest Limits In Fiber Optic Networks
Mark.JUK writes: Researchers at the University of California in San Diego have demonstrated a way of boosting transmissions over long distance fiber optic cables and removing crosstalk interference, which would mean no more need for expensive electronic regenerators (repeaters) to keep the signal stable. The result could be faster and cheaper networks, especially on long-distance international subsea cables. The feat was achieved by employing a frequency comb, which acts a bit like a concert conductor; the person responsible for tuning multiple instruments in an orchestra to the same pitch at the beginning of a concert. The comb was used to synchronize the frequency variations of the different streams of optical information (optical carriers) and thus compensate in advance for the crosstalk interference, which could also then be removed.
As a result the team were able to boost the power of their transmission some 20 fold and push data over a "record-breaking" 12,000km (7,400 miles) long fiber optic cable. The data was still intact at the other end and all of this was achieved without using repeaters and by only needing standard amplifiers.
As a result the team were able to boost the power of their transmission some 20 fold and push data over a "record-breaking" 12,000km (7,400 miles) long fiber optic cable. The data was still intact at the other end and all of this was achieved without using repeaters and by only needing standard amplifiers.
Actually, perhaps the article should perhaps correctly use the term 'concertmaster' (in orchestras, usually the first violin player), who also synchronizes the bow movements of the bowed instruments in the orchestra. Scanning the article linked, it seems most of what USCD is trying is a digital version of what is known in the analog world (particularly FM radio) as pre-emphasis. This principle has been known in the analog world for about 80 years, so it's kind of a shock that these techniques are not better known in the digital world. I think that this is actually perhaps a form of laziness, given the very high bandwidth and signal to noise ratios possible with digital that are not as much of an issue until very long distances are involved. I don't use that term loosely, as I've held an FCC commercial license for over 30 years, so basic engineering stuff like this coming out of a lab is a bit shocking. Perhaps a bit of old school techniques need to be applied to modern communications to get more out of them.
Interesting article but very light on details. I would love to read the actual paper but looks like it was published in Science. The actual press release here: http://ucsdnews.ucsd.edu/press... has slightly more info than the linked article. This link to the PDF from August 2014 with the theoretical basis is free: http://ieeexplore.ieee.org/sta... It looks like they are boosting WDM signals so this would work with existing long-range infrastructure.
To help understand the scale, the cable length is approximately the diameter of the earth (12742 km).
It is also 25-50% longer than the undersea hop for the longest cable paths (NY to London, LA to Sydney, San Francisco to Tokyo, Sao Palo to Gibraltar, etc.). This has the potential to allow electronics to stay on land, where they are easily maintainable and upgradable and with easy access to electricity.
Interesting development, indeed.
It's crosstalk between wavelengths on the same strand. Read up about CWDM and DWDM.