SDN initiatives are gathering steam on many fronts. Taking both the Grand Award and SDN category award in this year's Best of Interop product competition, the Open Daylight Project’s Hydrogen shows that even though software-defined networking is in its infancy, it’s also on many people’s minds as the obvious evolution of highly interoperable networking. And this is where we start feeling a bit of tension when contemplating how SDN and wireless networking might hook up.
If developed right, SDN will be the poster child for interoperability. However, the WLAN industry is in many ways the antithesis of interoperability. There is zero compatibility across each vendor’s product sets. I can’t connect a Ruckus wireless access point to a Cisco controller any more than I could hook up my television to the coffee pot. Unlike with Ethernet switches and routers, there is absolutely no inter-vendor mixing of wireless infrastructure products in a space where vendor lock has become the rule.
Despite the hyper-proprietary nature of WLAN systems, wireless vendors are starting to talk up SDN. But the messages are bound to be confusing and lead to questions on whether SDN will simply be leveraged within the confines of any one vendor’s own technical borders or whether software-defined networking could actually allow WLAN customers to somehow use components from multiple vendors in an end-to-end SDN-controlled environment. This presentation from Aruba Networks feels pretty Aruba-centric, whereas Meru Networks' approach with SDN and Wi-Fi gives hope that maybe SDN can provide far-ranging benefit to wireless customers beyond simply being a replacement for proprietary “magic in the middle.”
If you go back to the early days of controller-based WLAN, protocols like CAPWAP and LWAPP were briefly touted as a wonderful new hope for standards-based wireless systems interoperability. That didn’t last very long as each vendor immediately started building systems that were more akin to proprietary PBX telephone systems than they were to standards based LANs, and that has been the model ever since. I personally can’t envision any WLAN vendor giving up that level of control, and would be very surprised to see either “open” or “daylight” really come to our Wi-Fi systems.
[Get advice on how to pick out the right WLAN for your business in "How To Shop For A WLAN: 7 Steps."]
At the same time, wireless vendors are at a thorny juncture. With many market leaders also providing Ethernet switches and back-end role-based security and client onboarding and control mechanisms, ignoring SDN’s advent would be a poor strategy. But truly honoring the spirit of SDN would unravel customer reliance on single-vendor solutions. I’m guessing we’ll see lots of fashionable mentions of SDN capabilities trickling out of the wireless industry, as anything less would create a sense of technical stagnation that vendors can ill-afford.
At the end of the day, however, SDN will likely just end up being some sort of vendor-by-vendor API framework once you leave the LAN and get into the WLAN. This wouldn’t be a bad thing, but it would put an ironic twist on the notion of openness.
What do you think will happen with SDN in the wireless space? Please add your thoughts in the comment section below.