The ROADM renaissance begins
Reconfigurable optical add-drop multiplexers (ROADMs) are widely deployed in long-haul and metro networks today, and provide flexibility through automated configuration of bandwidth at the photonic layer. However, incumbent technologies and architectures have imposed constraints on reconfigurability and automation, particularly at add/drop nodes.
In reality, ROADMs to date have partially automated optical networks. They automate express nodes along a route but still require physical connections and truck rolls at the end-points of those connections. For operators, ROADMs have been valuable, but also limited.
The Next-Gen Renaissance Begins
To solve the end point connectivity problem, a new generation of ROADMs adds colorless, directionless, and contentionless functionality:
- Colorless: Automates assignment of add/drop wavelength to any available physical port
- Directionless: Allows any wavelength to be routed to any direction served by a node
- Contentionless: Allows multiple copies of the same wavelength on a single add/drop structure (without partitioning restrictions)
A 4th next-gen function, flexible spectrum or flexible grid, provides future proofing for 400 Gbit/s and higher channel rates to come by accommodating many variations of channel spacing. In combination, the functions are called CDC-F, though operators may not always choose to use them all together.
A key near term benefit of CDC-F for operators is the opex savings delivered through “touchless activation” - not having to send technicians to physically pull connectors at add/drop sites. Full automation of the photonic layer also:
- simplifies planning and operations (opex savings)
- reduces human error (more opex)
- enables rapid provisioning of optical capacity (a services differentiator)
Near term capex savings result from combining photonic layer and large scale OTN layer switching in order to switch capacity at the most efficient layer possible. A recent Alcatel-Lucent Bell Labs network simulation study showed capex savings of 20-37% using Next Gen CDC-F ROADM and OTN switching when compared to OTN switching alone. Other network simulations studies we have seen also place combined ROADM/OTN capex savings within this range.
ROADMs Ready to Take Over the World
Much fuss was made about CDC-F ROADMs a couple of years ago, but some factors had to come together to make commercial deployments a reality. All of the components hardware required for ROADM sub-systems had to mature, including larger 1x20 WSS elements and new MxN switches for contentionless. The move to flex spectrum hardware required, for some components makers, a major technology change-over.
In addition to functioning hardware, the software control to automate the hardware needed to be developed, matured, and (in cases) standardized. In fact, in our early discussions around CDC-F ROADMs in 2010 and 2011, components suppliers often lamented that the hardware development was running well ahead of the software needed to control it. Among the critical functions performed by the control software are:
- optimal path computation with multi-layer interaction
- wavelength power control
- dynamic wavelength routing for restoration
- flex spectrum control
Now the software side has come up to speed with the hardware and CDC-F is poised for major deployments worldwide.
Today, any software control mention quickly leads to a discussion of software-defined networking (SDN) – but for CDC-F ROADMs, the 2 need not be the same. GMPLS-based software control can accomplish all of the benefits we’ve listed so far in this column, including the multi-layer benefits of OTN layer and DWDM layer interoperability. “SDN” can be used quite loosely – among vendors and service providers alike.
In the context of transport, Heavy Reading sees SDN as a key enabler for multi-layer and multi-vendor interoperability, particularly for integrating IP and optical layers. Note that the predominant trend among operators is physical integration of OTN switching with DWDM transport (single vendor), but once the IP layer is involved, multiple vendors will typically be involved.
Big operators have done interesting studies showing the benefits of IP and optical integration via SDN control, but this is a work in progress. The key points to understand are:
- a switched photonic layer is a prerequisite for SDN at layer 0
- CDC-F provides opex and capex benefits in advance of SDN
In summary, while the concepts of end-to-end automated photonic switching have been around for many years, it has taken time for both the hardware and the software to develop and mature. We believe that with the enabling technologies now ready for prime time, operators globally will be moving quickly to reap the benefits the industry has been heralding for so long.
TechZine wishes to thank Sterling Perrin, Senior Analyst at Heavy Reading, for this guest post. To contact the author or request additional information, please send an email to firstname.lastname@example.org.