Extending enterprise UC networks over Wi-Fi and cellular networks

By Peter Thornycroft, Office of the CTO, Aruba
Share Post

A few years ago we were very excited about Fixed-Mobile Convergence, integrating the corporate PBX with cellphones.  We don't hear the acronym any more - what happened to FMC?  Like many ideas in technology, it may have been a little ahead of its time, but I believe we are starting to see a strong echo, with the same underlying intent but standing on the shoulders of newer technology.

FMC came about to bridge the two islands of communications: devices – mostly desktop phones – connected to the PBX with a private numbering scheme, and the cellphones that the majority of corporate employees use for business with public phone numbers.  The key technologies at the higher protocol layers were number manipulation and call forwarding, and at the lower layers we had to get the phone onto the corporate WLAN with Wi-Fi and IP connectivity and a VoIP client.  When outside Wi-Fi coverage, FMC came down to coordinating call control between the PBX and a software client on the phone, but the voice connection over the cellular network was a standard voice call.

Looking back, this was a sizable parcel of new technology, and it took a good deal of engineering to make it work.  Looking forward, is there still a need?  I'd say yes.  We still have a gulf between the PBX/UC systems that corporations use for communications inside the firewall and the smartphones and tablets that many employees carry and work on.  The fundamental technology requirements are still directory services and universal multi-media connectivity - being able to address and call a colleague no matter which device or connection is in use at either end.

Two technologies are making this a lot easier.

The first is automation at the Wi-Fi layer.  All smartphones today have Wi-Fi interfaces.  Configuration is relatively easy, and with BYOD, enterprise employees are accustomed to configure the phone for the corporate WLAN, automatically connecting and authenticating whenever it sees the signal, without any user intervention.

We eagerly anticipate the Wi-Fi Alliance 'Hotspot 2.0' certification ('Passpoint' when it launches) which will allow even greater automation of this Wi-Fi connection, but it's already reliable and silent, adequate for our purposes.  We owe much to the impetus from cellular offload, which has been the strongest driver of these features:  now that we can rely on the phone to discover and connect to the corporate WLAN without user intervention, the whole FMC system becomes more robust.

The second important driver in enabling the next generation of FMC is the steady march of VoIP, and particularly SIP in public and private networks.  In early FMC systems, we used VoIP between the FMC server and the phone, but this required a custom client on the phone and often a SIP upgrade to the corporate PBX.  And whenever the phone was outside Wi-Fi coverage, it had to use cellular voice bearers with a completely different client stack, which in turn made handover from Wi-Fi to cellular a complex operation.  A number of trends in VoIP are helping us.  First, SIP clients are readily available on today's smartphones, either organic to the OS or as apps on the marketplace.  These clients work easily with enterprise UC servers, and with every passing year more of the corporate PBX/UC installed base becomes SIP-capable.  Most enterprises today could configure these existing features to replace deskphones with SIP-enabled smartphone clients, if they wished, without any special or proprietary protocols.  With today's smartphones you will get decent multimedia performance on a decent enterprise WLAN.

The third area we see VoIP-friendly improvement is the cellular network.  Cellular data continually achieves higher rates, more reliability and wider geographic coverage.  I'd suggest that it crossed the threshold where one can rely on VoIP over cellular data some time in 2011 – with a reliable 120kbps up- and down-stream you can get a solid VoIP connection.  A number of alternative carriers are already exploiting this, and the Mobile World Congress this year was awash with 'over the top' service suggestions.  The key implication for FMC is that we can now see a future where the phone has a reliable IP connection whether in Wi-Fi or cellular coverage, and hence we can use a single VoIP client stack, making the switch between networks much easier.

There's an interesting short-term window where cellular voice becomes an island, as cellular operators don't generally have VoIP gateways (you can't make a call using your cellular number from a client on your PC).  But with my mobile SIP client I can reach my PBX or FMC server or SIP server over Wi-Fi or cellular data, and maybe I don't need to use that cellular phone number any more?  I'm hopeful that operators will see VoIP as an opportunity rather than a threat, dust off IMS and update it to run VoLTE and even offer Internet-connected calls via external VoIP gateways.  That's a little speculative, but there will certainly be disruption in the near future and operators will have to adapt.

To return to our theme, FMC is sleeping not dying.  It will return in several forms, from the updated traditional model described above to more proprietary architectures such as Microsoft Lync, as all UC developers are grappling with mobile devices, cellular and Wi-Fi.  Multi-modal communications is a given:  voice, video, IM, directory, presence etc.,  but all of that rests on the foundations outlined above.  When the mobile device moves smoothly between Wi-Fi and cellular networks, providing a reliable IP connection for higher-layer communications, the FMC system becomes far easier to implement.

Our mission at Aruba, of course, is to enable that solid Wi-Fi foundation by contributing to and implementing standards wherever possible - I'd point to the new Voice-Enterprise and forthcoming Hotspot 2.0 (Passpoint) certifications as important milestones - and building our own features where necessary to supplement those standards.  Multimedia-over-Wi-Fi-on-smartphones performance is pretty good already, but it's certainly going to improve.