Living on the Open ROADM: Can Open Source Disrupt Metro Networks?
By Casimer DeCusatis | Posted: 9 August 2016 3:18:36 PM
Twitter | LinkedIn | Facebook | Reddit
There’s a long list of industries that have been disrupted by new technology and open source software in recent years. Anyone who’s stayed at an AirBnB, shared a ride with Lyft, or read the news from a website instead of a traditional newspaper can attest to the dramatic impact which new technologies can have on an existing market. The optical communications world is no exception, and one of the biggest news stories to come out of OFC 2016 was AT&T’s announcements about Open ROADM technology.
Optical communication remains a hot topic in the industry, and plays an important role in the backbones interconnecting wireless access points. With data traffic on some wireless networks growing by more than 150,000% between 2007 and 2015, and no end currently in sight, we can expect major network service providers to invest in performance enhancing and cost saving upgrades for some time to come. The Reconfigurable Optical Add/Drop Multiplexer (ROADM) is a key component of any wavelength multiplexed network, and is used to adjust the amount of traffic on a particular wavelength or channel in the network. If dynamic bandwidth control is required, a software controlled ROADM can automatically sense and adjust optical network capacity, for example in response to bandwidth spikes or routing around failed components.
Impact of Disruptive Technologies
While software defined ROADMs are a powerful technology, some companies are turning to even more radical forms of innovation. However, the impact of disruptive new technologies is proving hard to predict and somewhat controversial. At OFC 2016, AT&T introduced the “Open ROADM” concept http://www.ofcconference.org/en-us/home/news-and-press/ofc-video-library/ofc-plenary-session-2016/ ), which attempts to bring the notion of open source design hardware deeper into the Optical Line Card (OLC) market. This effort was introduced as part of AT&T’s broader initiative in software defined networking (SDN); they have projected that over 75% of their network with be software-centric by the end of the decade. This move towards commoditization of optical hardware isn’t without its detractors; other presentations at OFC 2016 from Verizon question whether this approach makes sense for large incumbent carriers.
Benefits of SDN
Some proponents of SDN have argued that a move to centralized software control of the network inevitably leads to hardware commoditization, at least in the long term. But commodity hardware shouldn’t be the primary goal of an SDN network. Many presentations at OFC demonstrated the technical and performance benefits of SDN, such as higher capacity optical switching and interconnects as well as automation, analytics, and more. Prior research suggests that SDN also offers benefits such as faster provisioning times, better security, energy savings, and other benefits. Many of these benefits are being realized today using vendor proprietary versions of SDN, rather than a strictly open approach. This suggests that in the short term, SDN will find practical applications without impacting hardware costs, except in certain environments like hyper-converged data centers. A working draft specification has been published on the Open ROADM multi-source agreement website, and early supporters include Ciena, Fujitsu, and Nokia. Even if the Open ROADM proposal succeeds, the widespread adoption of multi-tier SDN in telecom networks doesn’t necessarily mean the end of high reliability hardware or a collapse in OLC prices.
Benefits of Open ROADM
There are many potential benefits to adopting an Open ROADM strategy. An open, programmable ROADM would enable service providers to choose from multiple component suppliers and use the best available technology at a price point they can more easily afford. The move away from vendor lock-in and towards open source hardware and software was a recurring theme at OFC this past year. Equipment suppliers will need to adapt to this changing world, and have greater incentive to compete and innovate in order to differentiate their offerings in an open marketplace. Clients and end users should benefit from improved technologies which provide lower latency, reduce network congestion, and accelerate the deployment of new interoperable technologies. A network using ROADMs from multiple vendors would allow wavelengths to be more easily moved to different paths, optimizing traffic flows and helping avoid overbuilding the network infrastructure. Dynamic reconfiguration of the optical layer helps automate maintenance and service activities, and facilitates multi-layer network management. In the future, migration to 5G technologies will likely involve deploying SDN and NFV capabilities at the cell site, driving a requirement for more resilient optical networks.
Do you think ROADMs are destined for a low cost, commoditized future? I’d like to know what you think, so drop me a line on Twitter (@Dr_Casimer).
Posted: 9 August 2016 by
| with 0 comments