Data signals that travel as laser pulses through an optical fiber are vulnerable to optical distortions resulting from interference between multiple signals of different wavelengths traveling through the same fiber.
These nonlinear wave interactions mean that data signals can degrade over long distances unless they are regularly regenerated along the waveform—that is, converted into electrical signals, subjected to computer analysis to remove distortions, and then converted back into optical signals. This process not only slows down data traffic but also accounts for the majority of the cost of building new optical network infrastructure.
Now, researchers at UC San Diego say they may have discovered a way to easily eliminate these distortions, reducing the need for constant and costly signal regeneration. The research team told the journal Science that this could lead to a two- to four-fold increase in both the amount of data a fiber can carry and the distance signals can travel before needing to be regenerated.
Lasers used to transmit signals through optical fibers typically vary in wavelength by hundredths of a percent as they operate. This variation is usually random, and the noise added to the data streams makes it virtually impossible to isolate and filter the distortions resulting from nonlinear wave interactions. Instead, researchers suggest making the variations in laser wavelengths predictable, rather than random.
Given that telecommunications networks typically use multiple lasers to generate the different wavelengths transmitted through optical fibers, researchers convert single-wavelength laser light into pulses of many different wavelengths. Experiments conducted with optical fibers over a thousand meters long showed that when the primary laser's wavelength fluctuated, its subsidiary pulses changed accordingly, matching quantities and making this simple variation easy to calculate. This, in turn, makes it easier to filter out distortions from nonlinear wave interactions.
Source: Charles Q. Choi
