SDN architectures

PON, 5G & the Case for Convergence

The Broadband Forum has a new project underway with Japanese operator NTT to optimize passive optical networks to support future 5G wireless services.

Like many other initiatives, the project is looking at how to virtualize PON network functions, but the Broadband Forum effort is specifically geared toward making PON infrastructure capable of providing 5G backhaul and fronthaul support. Fronthaul, in particular, is a time-critical application (TCA), and it brings with it requirements that today's PON networks aren't designed to meet.

The PON project -- known specifically as PON Abstraction Interface for TCAs -- is really an exercise in preparing for fixed mobile convergence. But it isn't the first time the communications industry has gone down this path.

"We tried doing fixed mobile convergence some time ago and were successful ... But did we get a whole bunch of fixed mobile convergence at that point? No, obviously not," says Broadband Forum CEO Robin Mersh. "And I would argue that the reason for that was timing. There really wasn't that overwhelming, compelling case."

With 5G, however, a business case for fixed mobile convergence may finally be emerging. High-capacity and low-latency wireless networks open the door to new revenue-generating services; services like those attached to connected cars, which will likely only be realized if and when 5G comes to market.

Assuming 5G is on its way, the Broadband Forum and its operator members want to make sure they're ready to handle it. But there are significant challenges based on the way that wired and wireless networks have evolved. The two types of networks were designed with different capabilities and constraints in mind. Wireless networks assume limited capacity and have a number of policy controls associated with managing bandwidth as a scarce resource. Wireless networks are also built to ensure roaming, whereas wired networks are less flexible in exchanging traffic.

It's not impossible to make one network behave more like the other, but there are costs involved.

And, as Mersh points out, "Why would you, say, add something that would be more costly in terms of network elements?" As in, why would you add functionality to a wireless or wireline network if that network doesn't need it to function today?

"You have to have business drivers really," says Mersh. "That's the bottom line. That's really where we get to is you say there have to be business drivers to be able to justify bringing in some of that new functionality."

Many in the industry are now betting that 5G will provide that justification.

You're invited to attend Light Reading's Big Communications Event – the ONE event that delivers fresh perspective on the rapid transformation of the telecom industry and the road ahead. We'll see you May 14-16 in Austin – communications service providers get in free!

In the PON project itself, there are two phases of work. The first will determine a policy for disaggregating PON functions that are time-critical and define the requirements for the interfaces needed to interact with those functions. The second will set interface specifications as API sets.

Mersh says the Broadband Forum is well aware that the work it's doing has to be applicable in environments that include both new virtual optical line terminals and standard hardware-based OLTs.

"There are some things you can do purely within the software of the current OLTs, so there are some things that you can just overlay," explains Mersh. "And certainly one of the things you would want to do, even bringing in a white box kind of solution, is you are going to have to coexist with those legacy OLTs."

The Broadband Forum is also in communication with other standards bodies working on PON virtualization and wants to coordinate when and where appropriate.

"We do know the other projects obviously that are going to be important out there. You can say for CORD or OpenNFV [that it's important to make] sure the people on the ITU know exactly what you're doing. Obviously they have some responsibility in the PON area. Same thing with the IEEE," notes Mersh. "So we can certainly make sure that everyone knows exactly what is being worked on. But I can't help but sort of think the market will really tell you."

In other words, strategic direction and open communication between agencies working on related projects are both important, but ultimately the technology will go where the market needs it to go.

Meanwhile, Mersh says the PON project has generated a great deal of interest and is directly responsible for attracting new members to the Broadband Forum. Those members have one goal: figuring out how to make sure their fixed networks can be a part of next-generation wireless service delivery.

— Mari Silbey, Senior Editor, Cable/Video, Light Reading

Sign In