Optical transmission was first used in communications networks because optical fiber was able to carry far larger quantities of data across much longer distances with much lower power consumption than copper wire systems.

An optical cable uses a large number of individual optical fibers as communication pathways. Once data finds its way onto a fiber, it is transmitted at the speed of light. However, moving data from one fiber to another requires time to establish the new connection. 

Actual optical switching was finally performed when microelectromechanical systems (MEMS), liquid-crystal, bubble, and other types of optical switches were developed.  This led not only to lower latency, but lower costs, since the high-speed electronics required for OEO conversion were expensive. 

Thanks to MEMS, optical switching then had lower latency than packet switching plus it had the added benefits of higher capacity, lower power consumption, and lower capital and operating expenses than before. However, optical switching still had a big drawback: it took a long time to reconfigure a transmission path. This limitation made it suitable only for transmitting large amounts of data along lasting connections.

However, with high-capacity, low-power optical networks and switches, you could implement a new kind of topology. It was called spine-leaf topology, or sometimes leaf & spine.

Ericsson_Cloud_optical_switching_infrastructure_spineleaf.png

Spine-leaf technologies relied on optical switches to handle long-lasting transmissions of data. Optical switches could perform those types of transmissions faster and with lower power, but just as important, they kept those long-lasting transmissions from interrupting the flow of data through the spines. As a result, smaller bursts of data could be transmitted at a faster pace by the rest of the spine-leaf network. 

​Were spine-leaf topologies and the advantages of optical switching enough to satisfy the needs of datacenters?  Unfortunately, no.  Even though using optical switching was cheaper than doing OEO conversions, optical switches were still expensive for datacenters.  And they were bulky.  Datacenters did not have the vast amounts of network traffic that made optical switches cost effective for telecom operators.

However, two recent developments are changing that equation. Explore this more in our paper on combining packet and optical switching:

Download the paper

About the Photograph

I took the photo of the eclectic storefronts of Miklabraut Street while facing toward the South East in Reykjavic in mid-April of 2016.


Cloud Infrastructure

Rick Ramsey

I started my high tech training as an avionics technician in the US Air Force. While studying Economics at UC Berkeley, I wrote reference manuals and developer guides for two artificial intelligence languages, ART and SYNTEL. At Sun Microsystems I wrote about hardware, software, and toolkits for developers and sysadmins, and published “All About Administering NIS+.” I served as information architect before joining BigAdmin, which morphed into the Systems Community at Oracle. I left Oracle in May of 2015, and now write for Ericsson.

Follow Rick Ramsey:
Rick Ramsey

Discussions