|
Boosting bandwidth
|
24/10/2008
|
| |
High speed transmission and broadband access are hot optical communication topics if this year’s European Conference on Optical Communication (ECOC) was anything to go by. On the exhibition floor at ECOC, held in Brussels in September, firms promoted 40Gbit/s components and emerging 100Gbit/s technologies, and detailed optical access schemes to come such as wavelength division multiplexing passive optical networks (WDM-PON).
One technology evident at ECOC, albeit with a far lower profile, was photonic integration. Optical integration is as old a concept as the integrated circuit but, unlike the semiconductor industry, has no Moore’s Law. The impact of integration on the optical industry has thus been muted, although there are sound technology and business reasons for this.
First, optical integration drivers differ greatly from those of the semiconductor industry. There is a limit to how much feature size shrinking makes sense. That’s because the wavelength of light defines the optimum dimensions – in microns – of an optical waveguide use to direct a signal.
The semiconductor industry has also long concentrated on one material – cmos – whereas optics uses several. Indium Phosphide (InP) is one, used for active components such as lasers and photodetectors, while planar lightwave circuits (PLCs) is another. Based on silica on silicon, a PLC is used to implement passive functions such as waveguides, filtering and switching.
InP itself can implement optical waveguides to create more complex optical circuits that combine active and passive components. But such monolithically integrated devices are relatively expensive and suffer chip yield issues with each added function.
|
| |
Author
Roy Rubenstein
|
| |
| |
|
This material is protected by Findlay Publications copyright 2009.
See Terms and Conditions.
One-off usage is permitted but bulk copying is not.
For multiple copies contact the sales team.
|
| |
|
| Email this article |
| |
|
|
|
|