Telecoms.com periodically invites expert third parties to share their views on the industry’s most pressing issues. In this piece Alexander Seyf, Partner at Sytel Reply, looks at some successful SDN implementations and stresses the importance of open source.
The rise of programmable networks is transforming the way telcos operate. Many service providers are now successfully implementing SDN and NFV into their architectures – both of which are enabling them to accelerate their time to market and reduce costs.
Verizon has been laying the foundations for its next-generation network for several years. Last year, it announced a new SDN-based architecture designed to introduce new operational efficiencies and enable the rapid and flexible delivery of services to its customers.
Meanwhile, AT&T has been putting some of its network under software control for the past few years, and it has ambitious plans for growing its SDN-enabled network. Using SDN, the carrier plans to transform over 75% of its network to software by the year 2020. This year alone, it will launch a new SDN service in 63 countries.
It is therefore no surprise that both SDN and NFV could eventually replace today’s carrier networks – most of which are built around proprietary hardware. Future networks will lead to greater interoperability, more innovation, and more flexible, cost‐effective solutions. The key point here is interoperability. Without it, there will be increased vendor lock-in, and this will make network management too complex. A single company can’t have the intelligence and technologies to manage an entire network, and this calls for collaboration.
The original aspiration for a single SDN controller for an entire network has now made way for a federated approach. If a network is compliant with industry-wide SDN standards, it can seamlessly be controlled by multiple SDN controllers – and therefore multiple vendors. And there is demand for this too. A recent study cited by Brocade suggested that 90% of network managers saw a truly open SDN controller with multiple supplier support as essential to SDN deployment plans.
Key open source initiatives
Ultimately, successful SDN strategies will be based on interoperable multi-vendor ecosystems with open source technologies or standardised protocols. We are already seeing some of the largest operators championing interoperability and openness, with the development of multiple SDN standards in different areas.
AT&T as we know is an advocate of SDN, but it is also offering to work with open source communities. Its Domain 2.0 programme, which launched in February 2014, seeks to deploy an open source based software framework suitable for large carriers, which AT&T and others can use and contribute to freely. A large number of vendors are signed up to the programme including Ericsson, Affirmed Networks, Metaswitch Networks, Amdocs, Juniper Networks and, most recently, Nokia.
AT&T’s Enhanced Control, Orchestration, Management and Policy (ECOMP) architecture is likely to follow suit after reports that the carrier is seeking input from developers on whether to make it open source.
What is crucial to the success of any open source initiative is that cooperation should take the place of competition. Telcos need to be able to work together in the development of open source software used to provide SDN and NFV services. Put simply, it makes business sense for them to collaborate. They all share the same needs and desires on the technologies that will make future networking work, yet each will have different capabilities to offer that, combined, will be for the public good.
This level of collaboration is evidenced by the Open Network Operating System (ONOS) project, which serves to accelerate the adoption of open source SDN and NFV solutions. The community is made up of over 50 partners and collaborators, which include rivals AT&T and Verizon, plus China Unicom, NTT Communications and SK Telecom.
And then we have the New IP Agency (NIA), which formally launched in January 2016 with significant support from the carrier industry. As we move to future networks built on open standards, an independent entity like NIA can provide the facilities for vendor interoperability testing.
As stated on Virtuapedia, every test is intended to provide tangible and objective information about the progress and adoption of virtualised solutions. The results of these tests support service providers’ adoption of virtualised solutions by minimising the need for service providers’ own lab tests. The NIA’s results will also provide the telco industry with a matrix of what works with what, and who works with whom.
LSO: a contender to open standards?
There are now dozens of open standards projects including ‘official’ organisations and single-vendor initiatives – and they are all spearheading the role of interoperability in future networks. But this raises the question as to whether there are too many standards.
Some experts are going as far as saying that there is a better solution for creating faster and more agile networks, and what is needed is a standardised service orchestration solution such as the Lifecycle Service Orchestration (LSO). Industry analyst Alan Zeichick argues telcos should ‘Forget SDN and NFV: It’s all about LSO’. Meanwhile, Tom Nolle, President of CIMI Corp, explains why the best approach to SDN and NFV isn’t from ETSI or ‘Open-Something’, but from the MEF (Metro Ethernet Forum) and its LSO capabilities.
As with the open source initiatives outlined above, LSO provides better multi-vendor interoperability of services but network operators need more according to research. While open standards enable the creation of technologies that can work with each other in a network, each technology can be implemented individually but it lacks orchestration.
LSO, on the other hand, is a set of specifications defined by the MEF, which streamlines and automates the service lifecycle for coordinated management and control of end-to-end connectivity services over one or more network service domains.
In combination with SDN and NFV, LSO is designed to enable the Third Network – a vision that delivers internet-like agility and ubiquity with CE 2.0-like performance and security. According to MEF, the orchestration capabilities are the backbone of the lifecycle services provided by future networks. They will dramatically decrease the time to establish and modify the characteristics of the end to end service as well as assure the overall service quality and security guarantees for these services.
In summary, the fate of future networks is still unknown. SDN and NFV are still largely untested, while LSO is relatively new. What we do know is the telco industry needs to work together to create a new kind of network that is easy to provision and automate for all.
Alexander Seyf is Partner of Sytel Reply. His roles have provided substantial personal exposure to the traditional and evolving technologies and business drivers within the Telecom and IT industries. He has also managed, developed and delivered many major bids and negotiated / managed relationships with suppliers and alliance partners.