The mania and hype around software-defined networks is at an all time high. It reminds me of Wrestlemania IV when Macho Man Randy Savage partnered with Hulk Hogan and the Mega Powers united and achieved a level of hype never seen before in the WWF. Unlike SDNs though, The Savage/Hogan tandem actually delivered and both went on to win several WWF belts (although Savage and Miss Elizabeth are both dead now...).
The reason I say this about hype is that most companies I talk to couldn't give two hoots about SDN. Additionally, most application developers really have no concept of how to leverage an SDN to build more intelligent applications.
There are a couple of exceptions, though. Most of the service providers I've interviewed do have an interest in SDNs, particularly to enable virtual services and manage the network more easily. Also, the only applications right now that can leverage SDNs are ones that rely heavily on the network, like UC services.
This intersection of service providers and UC is exactly where Juniper and Sonus have aimed a joint solution. Earlier this week, the two companies announced a strategic alliance where they would come to market with joint SDN solutions for video and other unified communication services.
Through the agreement, the Sonus 5000 series session border controller (SBC) will be integrated with Juniper's MX series router and SRC policy engine to allow analytic information from the application/session layer to be pooled together with network intelligence, allowing service providers to intelligently deal with the rising amount of UC traffic, particularly video, without having to over-provision the network. Over-provisioning, of course, is the most common way of dealing with the seemingly endless amount of video traffic flowing across service provider networks today.
The combined solution will enable network operators to simplify network design and ongoing management, which ultimately lowers the overall total cost of ownership. The application-specific intelligence from the Sonus SBC, combined with Juniper's knowledge of traffic flows at the network layer, gives service providers the ability to offer levels of quality of service that were not possible without a significant amount of custom coding. Additionally the combination of the two gives the ability to enforce those dreaded service level agreements in real time instead of the significant lag seen today. This should be one of the key value propositions of SDNs today. Much of what SDN promises can be done today, but it could be much easier and faster by leveraging SDN principles.
Additionally, the two companies are developing integration points between the Sonus SBCs and Juniper's JunosV App Engine, which could lead to the Sonus SBC being offered as a virtual service within a Juniper MX router. This would be an interesting move for Juniper, as it has a virtual SBC today with the assets from Kagoor, but Sonus would give Juniper significantly better technology.
The tag-team solution is well timed, as the growth in UC is likely to be in cloud-delivered services. While I don't think the premise-based solutions are going away, the cloud solutions have reached near feature parity and are viable options for companies, particularly those with limited IT skills or budgets. Also, cloud video services are increasingly popular today, as leveraging the cloud is the only scalable, cost effective way of implementing B2B video services.
While this should be music to the service providers' ears, it does present a bit of a "be careful what you ask for scenario". Cisco's Visual Network Index has mobile data growing 13-fold from 2012 to 2017, with about two-thirds of that traffic being video. I actually think these numbers are conservative, as we'll see richer, more bandwidth-intensive forms of video traversing the networks as the technology gets better and better.
In theory, the Juniper-Sonus solution gives joint customers a level of control where service providers can get aggressive with rolling out UC services without the risk of over-burdening the network and creating poor service for all. I believe cloud UC services will be one of the most hotly contested markets over the next five to ten years, and those service providers that offer the best quality and can actually meet SLAs will grab the majority of the share. New services and delivery models require new ways of managing the network, and that's what an SDN should deliver.
Even though the partnership between Sonus and Juniper doesn't solve all the world's problems like so many of the SDN vendors claim to, it is an actual use case targeted towards a specific audience. The joint solution is aimed at giving the service providers the capability to monitor and adjust policies for UC applications, specifically video. Considering the complexity of doing this today, it's a great, practical use case of software defined networks. I don't believe the solution is available yet, but it is being demonstrated in Juniper's Innovation Lab in New Jersey. If the two companies can actually deliver on the vision promised, they'll not only hit the target, but should hit the bullseye.