Rebuilding Optical Networks for Packets Optical switching has become an important pillar in telecom networks, but its development bears very little resemblance to early visions of the how the technology would be deployed and used. In optical crossconnects, for instance, demand for all-optical products did not materialize, but crossconnects with optical-to-electrical-to-optical (OEO) switch fabrics – which convert light into electricity for processing before reconverting it to light for transmission – gained a ready following. Unlike all-optical crossconnects, these OEO products were able to groom traffic at lower speeds such as 155 Mbit/s, which is exactly what the market required.

Meanwhile, some all-optical fabric technologies proved important, but the end devices are very different from the Xros and LambdaRouter concepts promoted in the early part of this decade. All-optical switching technologies – particularly micro-electro-mechanical systems (MEMS) switching fabrics – are being sold in volumes today. However, they are deployed as the foundation of reconfigurable optical add-drop multiplexer (ROADM) subsystems, not as large optical crossconnects. Often, these subsystems are integrated into wavelength division multiplexing (WDM) transport equipment, particularly in metro/regional WDM systems. They are small in size and low in port counts (and much lower in cost).

As we look into the future, optical switching needs to evolve again, and this evolution is now underway. At the highest level, the changes are driven by the need to re-architect telecom networks for the transition from TDM to packets – a move that is just starting to happen. For the transport network, this means a change from Sonet/SDH transport to packet transport using Ethernet and IP protocols. For the switching network, this means a transition from circuit-switched Sonet/SDH to switching built for packet protocols – ultimately switching packets themselves optically.

Of increasing importance in optical switching are Ethernet switching and optical transport network (OTN) switching, both of which are built for packet traffic. Also critical are developments in the optical control plane, including: the International Telecommunication Union's (ITU) Automatically Switched Optical Network (ASON) recommendations; the Internet Engineering Task Force's (IETF) Generalized Multiprotocol Label Switching (GMPLS) standards; and the Optical Internetworking Forum's (OIF) implementation agreements for control plane interoperability.

An important difference moving forward is that the optical switching technologies and standards will not reside on a single network element (i.e., the optical crossconnect), but will reside in different network elements located in different parts of the network and operating at different layers of the Open Systems Interconnection (OSI) stack (i.e., Layers 0, 1, 2, and 3). This more distributed role for the optical switching function is a positive development for the industry. There are more vendors and more minds at work tackling the switching problem. It does, however, make the switching world more complex and bring interworking and interoperability to the forefront.

The Optical Switching Revival: Rebuilding the Optical Network for Packets offers a detailed investigation into the future of optical switching, including in-depth analysis of the following product and technology areas:

			Optical control plane standards and standards organizations, including ITU ASON, IETF GMPLS, and OIF
			Optical crossconnects
			OTN switching
			Photonic switching with ROADMs
			Next-generation ROADM technologies, including colorless ROADMs, directionless ROADMs, and "digital" ROADMs
			Converged packet-optical transport systems (P-OTS)
			Optical packet switching (OPS)
			Optical burst switching (OBS) and especially ring-based OBS