Doubling the capacity of our communications networks
Currently around 95% of all the world’s long-distance communications travel through hair-thin glass optical fibres. These fibres are contained in bundles that run across continents and under the world’s oceans.
Thirty years ago when the first fibre networks were installed, their primary use was for the transmission of digitised telephone calls and faxes, but today, things are different. An exponentially growing demand for web services means that the bulk of fibre optic traffic comprises data transfer between different computers over the internet. To cope with this expansion in demand, engineers transmit several signals at once down the same fibre using different wavelengths of light. In effect, this has increased the capacity of existing networks by several orders of magnitude, without the need to lay any more fibre.
Because the fibres in very long distance transmission systems can only operate over a relatively narrow spectral range, there are physical limits to how many different wavelengths can be used simultaneously. As a result, we are now reaching a situation where even multi-wavelength transmission will not cope with increasing demand. The limits on the capacity of existing multi-wavelength fibres are expected to be reached within less than a decade.
Just because you can’t add more channels by adding more colours doesn’t mean that there is no potential for improvement though. Two ANU researchers, Professor John Love and Nick Riesen are currently working on novel systems for injecting multiple signals into fibres separated by what are known as modes.
A mode essentially describes the possible ways a beam of light can move along a fibre by reflecting off the walls. Existing optical fibres have a very small central light-guiding core region, typically about 10 microns in diameter and are designed to carry just one mode with negligible distortion and attenuation.
However if the fibres were able to support two or more modes simultaneously and the modes were independent of one another, this would mean that in principle the overall capacity of the fibre could be easily doubled, trebled, etc. by adding more modes.
One of the major obstacles in achieving this in practice is getting the different modes into the fibre and separating them at the end without mixing them all up in the process. It is this area that forms the focus of the work at ANU.
Using specially designed couplers, the researchers have shown that it’s possible to bring two or three single mode fibres together and inject each of their signals into separate modes within a single fibre. At the other end, a similar device extracts the modes and channels them into separate detectors.
An ingenious feature of this new coupler is that it avoids signal dropouts caused by interference nulling that would occur if existing couplers were employed in this task.
How many of these separate modes it’s possible to squeeze into a single fibre depends on its structure. Existing single mode fibres can accommodate only one. If they were just 10% bigger, that number would be two and 15% bigger would enable several modes to propagate.
Whilst it may not be practical to increase the diameter of existing fibres, it is possible to play tricks with the physics. Reducing the wavelength of the light by 15% achieves essentially the same thing as increasing the size of the fibre because the light itself now becomes 15% smaller.
For existing cross-city and urban transmission systems that mostly use single-mode fibres, the shorter wavelength and mode-selective couplers could provide a practical solution to increasing their capacity. And for most of us, that would mean faster internet.
Existing very long distance fibre systems however are a different story. They require amplification of the light signals at intervals of about 50 km’s. This amplification is achieved using specially incorporated impurities within the fibre and is powered by a second laser beam travelling down the outside. Unfortunately this system simply won’t work at shorter wavelengths.
This means the “few-mode” transmission option can only work in future long distance systems that are laid with physically larger cores. But if the technology proves its worth on the short networks, there would be a great incentive to lay larger cores in future long distance fibres.