While recent blog posts on low-latency networks were focused on speeds and feeds of data transport between two given locations the underlying physics were always considered as a constant that cannot be changed. It is widely accepted that fiber optic transport is the fastest and most efficient way to transfer information in networks and the speed of light seems to be the natural barrier here. To break this natural barrier, things like WARP drives or usage of wormholes could be the next alternative but unfortunately research has not reached a status yet that would make such things seem realistic (see NASA - Status of "Warp Drive"). But before we start thinking about speeds faster than speed of light we could actually still improve the speed we are using today for fiber optic transmission.

The maximum speed of light is defined in vacuum and fiber optic cables as we know them today are not transmitting the data in a vacuum but in a fiber core made out of silica. Unfortunately the silica refraction index, which defines the speed of light in that medium, is somewhere around 1.45 meaning that the light travels by a factor of 1.45 slower than it would do in a vacuum (200.000 km/s instead of 300.000km/s). So we are still far away from the theoretical maximum and there are ways to improve this.

A typical fiber consists of two main components: the core of the fiber and the cladding of the fiber. The actual light signal travels inside the core and in order to keep it inside this core it is reflected at the transition layer between core and cladding as it travels along the fiber cable. This total reflection can only happen if the refraction index of the core is higher than the one of the cladding. The term refraction index describes the propagation speed of light in a given medium, where an index of 1 stands for the maximum speed which is achievable in a vacuum only. (I'll spare you all the other underlying physics). Since 1 is the lowest achievable index value and since for a reflection of light the refraction index of the core needs to be higher than the one of the cladding the desired refraction index of 1 (meaning the light would travel at its maximum speed) is not achievable within the core of cables as we know them today.

But in theory there is a way around this: Special fibers that contain holes in the fiber cladding and the core which act as waveguides could be designed so that the light signal would travel only inside those waveguides without the need for a special refraction index. The speed of light in those air-filled wave guides would be very close to the theoretical maximum.

So by using those so-called hollow core or band gap fibers one could actually increase the speed of light inside the fiber cables by 50% thus reducing the link latency by 30% which for a 1000km link could be several milliseconds.

So much for the theory - unfortunately those fibers only exist for small distances within lab environments today and also problems like splicing, bending and termination need to be solved. So we might come back to the idea of WARP drives instead…

Read more on low-latency optical transport here.