4 Developments and future directions in optical-fibre communications
Additional material for this unit, by David Chapman, January 2005
The start of optical-fibre communication is generally identified with a paper published in 1966 (Kao and Hockham, 1966). It was not until about ten years later that it was commercially viable, but from then on there was more or less continuous development, with substantial research effort taking place both in industry and universities.
Innovation continues today, and this additional material introduces some of the fields of development that are being pursued now.
A good overview of the history of optical-fibre communications in the UK can be found in a paper by Professor Midwinter of University College London, published in the ‘Millennium Issue’ of the IEEE Journal on Selected Topics in Quantum Electronics (Midwinter, 2000).
Coherent optical communications: a cul-de-sac?
In the 1980s, a lot of research effort was expended on developing a technique known as ‘coherent detection’ of optical signals. The technique involved applying to optical signals a method that is widely used in the detection of radio signals, but that is much more difficult to do with optical signals because of the very much higher frequency of light (of the order of 10 Hz) compared to radio (under 10 Hz). Coherent detection provides a way of building receivers with very high sensitivity, and might also be used to allow closer channel spacing in wavelength division multiplexed signals.
Previously, I wrote:
At the time of writing (1988), although coherent detection has been demonstrated in many impressive experiments, it still requires further development before it will be suitable for use in commercial optical systems
Full exploitation of the capabilities of optical fibre can be achieved by the use of … coherent detection.
As I write now, 18 years later, commercial deployment of coherent detection has still not taken place, and the research effort has tailed off. To a large extent, interest in coherent optical techniques waned with the development of erbium-doped fibre amplifiers (EDFAs). The main attraction, in the short term at least, of coherent detection was the very high receive sensitivity that it gave, which in turn was desirable because it allowed longer transmission distances between repeaters. Optical amplifiers provided the same benefit (long repeater spacing), and it turned out that EDFAs were much simpler to use than coherent optical detection. The use of coherent techniques in wavelength division multiplexing might still be valuable one day, but for the present other, simpler, technologies are sufficient and coherent optical techniques looks like a research avenue that turned into a cul-de-sac.