Disaggregation, open networking and SDN control have become key themes when discussing the future of optical networks at industry events. All of these interrelated topics were certainly popular at this year’s Big Communications Event in Austin, Texas, where I joined Axel Clauberg of Deutsche Telekom and Huawei’s Sean Long for a panel debate titled Optical Networks in the New Age of “Open”.
Discussions of open optical networking regularly cover the many benefits it promises. We often hear how it provides significant financial savings in terms of both CapEx and OpEx. It also offers freedom from vendor lock-in and the opportunity to choose best-in-breed components, as well as simplified interoperability between equipment from different suppliers. Multi-layer SDN-based network resource coordination is another frequently discussed benefit.
But our debate was about the real future of “openness” and what the impact on the optical transport world is likely to be from the new open groups, standards and concepts that are currently emerging throughout telecom.
Two key applications are internet content providers (ICPs) deploying hyper-scale data center interconnect (DCI) networks and IP/optical integration with network operators. First live network deployments are well under way for both of these use cases.
For ICPs, the quickest way of reducing costs and reaping the benefits of new technology is through disaggregation and openness of optical network components in DCI applications. With their software-centric nature, ICPs find the task of integrating new components fairly straightforward. They also benefit from their relatively simple optical metro network architecture, which tends to be essentially point-to-point and medium-reach, stretching up to about 80 kilometers.
In the telecommunications sector, the focus is on integrating IP and optical networking to reduce interfaces and multi-layer coordination with the aim of saving equipment costs and increasing network utilization. The point-to-point nature of links between core routers also lends itself well to this scenario.
However, the effectiveness of an open approach is highly dependent on a large number of vendors adopting it. So far, it remains to be seen how committed the industry as a whole will be to supporting open strategies.
Up until now, progress toward this goal has been slow. This is partly due to continuing rapid innovation which is enhancing optical network technology. There’s also a lot of emphasis on optimizing cost and performance metrics, which is easier to achieve by tuning the entire system in contrast to looking at decomposed elements.
But progress has been made. And it’s not just the vendors who are driving it. Operators are also playing a significant role. A good example of this would be the Telecom Infra Project (TIP) led by Facebook, which involves several key players in the industry making a concerted attempt to push openness into optical network solutions.
One thing that the panel at BCE agreed on was that we’ll see more openness in optical networks going forward. Although, what this will mean for multi-vendor deployments is still far from certain. Ultimately, time will tell how openness in the network will truly manifest itself. But, over the next five years, it’s likely that the real value will only be exploited with some specific applications such as hyper-scale DCI and IP/optical multi-layer coordination.