While XGS-PON (10G PON) has caught on as the next evolution for fiber to the home (FTTH), service providers and vendors are eying what will be required to meet future broadband needs for consumers and businesses.
David Smith, chief network officer at Lumos Fiber; Harj Ghuman, vice president of strategic products at Precision Optical Technologies, Inc.; Niki Kirschenmann, global product line manager at VIAVI Solutions; and Fernando Villarruel, chief architect of broadband at Ciena, came together at BTR Summit 2024 to discuss the future of 10G-PON and what lies beyond it.
Smith, who noted that Lumos was one of the first to offer 1 Gbps speeds to residential customers (using GPON at that time), said that from a service provider perspective, evolution depends on the access layer and customer needs.
“One of the things we evaluate as a service provider as we’re architecting our access layers,” said Smith, “is what we’re seeing of throughput, what we’re seeing of traffic behaviors; really, that’s driving us to go to the next evolution because the access layer just gets us to the house, but then how do we take care of our customers? Because we’ve seen, over the last ten or so years, 80% or more of devices at our house have been hardwired, to now, 90% or more of the devices are wireless. So, we can bring a multi-gig connection into the home, but what happens in the home is really the customer experience, and that’s the way we kind of look at it.”
He noted that Lumos does not plan to roll out more advanced architecture at the access layer soon.
“We do know,” he said, “that eventually we will go to another generation. Right now, we believe that’s probably going to be the 50G standard: that’s what we think our next jump will be. Kind of like where we skipped the XGPON standard to the XGS, we probably will skip the 25G to the 50G—that’s kind of where we envision ourselves at in the current architecture.”
Ghuman said that in his experience, operators, especially the smaller ones, want something that’s easy to install and configure, and this will influence the future of PON.
“So, traditionally,” he said, “PON cycles last for ten years. I think GPON lasted about ten years, XGS-PON started being deployed in the last few years, and there is still a lot of bandwidth left in XGS-PON, probably another three or four years. And beyond that, I think that we as a company, are kind of looking at what the ecosystem is. Currently, there is 25G-PON deploying by one or two vendors in small areas; if the cycle holds up and 50G comes along in the next few years, operators may deploy it now instead of waiting.”
Ghuman also reported that Precision is looking into extending the range and length of PON, which he said is traditionally limited to 20 kilometers.
“If you look at the failures,” he said, “most of the failures come from power supply failures and stuff like that. So, we’re developing solutions that can be used with an OLT and can be housed wherever the OLT is and can extend the length of fiber by PON to beyond 60 to 70 kilometers. I think that would be a pretty disruptive thing.”
Evolving tests and measurement
Kirschenmann said VIAVI is preparing to stay at pace with the advancements in PON technologies and new opportunities bring new problems to overcome.
“I wouldn’t say that it applies to everyone, but there is a certain percentage that refuses to test their construction build,” Kirschenmann said. “So, they run into the issue, not now, but maybe sometime in the future, that something is wrong, and they need to find the fault. This is where we try to educate: ‘You simply need to test.’”
Kirschenmann said that it’s important for field technicians to use the appropriate field testers to distinguish between different networks and to strengthen their overall processes.
“In a lot of cases,” he said, “the process is just weak or too complex. The measurement task is there but think about what the field technician needs to do: someone needs to create a work order, someone needs to deploy these work orders to the technician, and the technician carries out the task and they redeploy it back, and someone needs to review all of that. So, this is where we are trying to be on top of things, and our portfolio extends from lab solutions to field solutions.”
The next evolution and coexistence
Regarding the next evolution of connectivity, Villarruel used Chattanooga, Georgia, as an example of where other cities should aim.
“Chattanooga markets itself as a ‘Gig city’—that’s their thing,” he said. “And people move there, and they love the internet: the broadband is amazing. What’s interesting is they’ve built their city on 25G. If you’re brand new, and a lot of our cable broadband customers are new, building on the latest affordable solution is not a bad idea, right?”
Villarruel also said that, with PON, it doesn’t have to be a one-or-the-other situation.
“I know we’ve thought about PON as ‘there’s this rigid step, and then there’s this rigid step, and then you throw everything out,’” he said. “But the reality is—the way Ciena has looked at this, via the Tibit acquisition, which we did last year—if you have an Ethernet underlay, then you do what the Ethernet underlay allows you to do, and most service provider switches are 10G or 25G. So any port can be either a 10G or a 25G port, and if that’s your underlay for either the remote-fi, or the DAA, or something else, then you really have the capability to do a study and say, ‘What’s best for me?’ in this particular area or for this particular customer set and do 10G or 25G because the Ethernet underlay allows that.”
Villarruel said it’s essential to think about coexistence now to plan for PON technologies’ future.
“Imagine six years from now,” he said, “you’re a power customer in a 10G deployment, and you’re stuck with 10G because the rest of your neighbors don’t use it: you would like to have the capability for 25G if that’s possible. So, I think we have to maybe think about PON in a fresh way: not all of the rules that have been there in the past necessarily have to survive in the future.”
Smith said that, for Lumos, coexistence has long been a part of the plan.
“We’ve had BPON, GPON, XGS-PON out there,” he said. “Obviously, BPON is gone from our network, but we still have GPON technology out there, and so we have deployed coexistence devices to make it where we’re able to offer across that same splitter, whether its GPON or XGS-PON.”
He said there is, however, a lot of work required in back-office systems to manage records around different functionalities.
“One of the things that we have out there now with some of our partners,” he said, “they might refer to them as combo cards—basically, you take out the coexistence device that you would put in front of your splitter, and now it’s at the PON port itself so the PON port can do GPON or XGS, and it kind of eliminates another failure point. That’s a great thing about this: with the amount of attention that the fiber-to-the-premise industry has gotten, the evolution of it has picked up a whole lot quicker than we saw from BPON to GPON to XGS-PON.”
For testers, the challenges and opportunities lie in this rapid evolution.
“The development of the GPON standard took a few years,” said Kirschenmann, “until all agreed that ‘this is what we want’ and ‘this is what we can work on.’ And then the development stages of XGS-PON, then 25G-PON, 25GS-PON, 50G; the time is way shorter, and people need to react super-fast on this. And in all of these standards, there are recommendations on the limits for an optical power that should be received at the customer’s premise.”
Kirschenmann said it’s important to have tools that, when customers input their specific architecture details, they can receive pass/fail for the power level received. This, he says, simplifies the job of the technician, who has little time to determine their instrument settings.
“The coexistence happens at the length layer through an Ethernet underlay through an Ethernet switch,” Villarruel added, “so that’s easy to do. But the reality on the ODN is that you’re just managing wavelengths, which is a blessing, let’s be honest. In the past, in DOCSIS, if we had to go from one technology state to another, there was a lot of fancy stuff going on, but managing wavelengths is something that operators have done well for a long time.”
MDUs
Villarruel said that when it comes to MDUs, being the first to move on an opportunity is pivotal.
“I think that the flexibility of having either the ONU at the base of the MDU and then using the infrastructure there or having an OLT at the bottom and distributing fiber within the ONU—that is another variation—is very powerful. But I do think these bids come fast and strong, and you must get up front,” he said.
Smith says operations around MDUs are market-dependent.
“In the areas that Lumos serves,” he said, “we don’t deal with too many high-rises or 20 to 30-story building MDU complexes; our MDU complexes are usually three to four stories tall, somewhere around 24 to 48 units per building. Our golden standard that we have in place is to get the fiber into the unit if we can because we know technology can change, but fiber has a long life, so if we just have to change out the end units, that’s great.”
The strategy also depends on the arrangements the MDU has with other providers.
“Is it a bolt-type of arrangement,” Smith said, “to where we light up all the units all at one time, and the whole complex has our service: we cover the pools and common areas with wi-fi, so we basically make this ubiquitous coverage of our internet there, versus a competitive MDU perspective.
When it’s a competitive environment, ease of customer access is Lumos’s strategy.
Kirschenmann touched on the importance of verifying the network connection in MDUs, with upstream speed taking on more importance in some areas.
“When we talk about different areas where we have these multi-dwelling units,” he explained, “we say ‘okay, how do we identify if the user receives what they should, and more importantly do the ONTs that are provided to the end user directly work in the network?’ So, we talk about this coexistence, and the downstream wavelength can easily coexist, but there are certain challenges with the upstream; but if you have this scenario and you want to verify the network or the connection, you need to know what the upstream is, so this becomes increasingly important.”
Ease of installation
Ghuman pointed out that ease of installation opens opportunities in the MDU space.
“When you’re deploying PON to the MDU,” he said, “essentially, you just need to bring one fiber to the MDU. Yeah, internally, you must put fiber in each unit space, but that’s a lot easier than installing in a subdivision, a ground fill especially, where you have to have fiber to every home. Normally, I see operators putting an OLT in the base of the building, or they can put it in a central office and just have splitters in the basement or in the building.”
He also said that upgrades are easier in MDUs.
“When we talk about upgrades or coexistence,” he said, “once you have fiber to the home, say it’s currently GPON and you want to go to XGS-PON, all you have to do is change the ends of the network, right? You change the OLT at the home and at the central office.”
Adding to Ghuman’s point on upgradability, Villarruel added that the emergence of pluggable ONUs is adding to the ease.
“You get fiber to the home,” he said, “then you change the endpoints. There are really three modes: you can either have a home gateway that has the ONU in it—basically a router for the home that has the distribution, the wi-fi, and all that stuff; you can have a two box solution, where you have a separate ONT and then a separate home gateway and then if you ever upgrade from one technology to another, you can keep the gateway, and just change the ONU. But what’s interesting now is we’re seeing a lot of home gateways that accept pluggable ONUs, which means the simplification is even easier. Now you’re talking about literally just taking out a plug and putting in a new one; the security measures between one box and the other get simplified.”
Looking forward
The session ended with some positive outlooks on the future of the industry.
“Let’s not, as a community, approach what’s beyond 10G Pon with any fear,” said Villarruel. “There are very smart people working very hard to make sure these transitions are digestible and easy. We’ve been through some hard transitions in our previous lives, with DOCSIS and DSL. A lot of the lessons there are things that folks are keeping in mind when it comes to PON transitions, and I think we’re doing a pretty job at it so far.”
“We’re in an exciting time in our space,” added Smith. “The technology is evolving faster than it has over the last ten-plus years, and it’s just going to get better.”
For more news, products, and technical profiles in broadband cable and telecommunications technology, subscribe to BTR's newsletter and follow us on LinkedIn, Twitter, and Facebook.
Hayden Beeson
Hayden Beeson is a writer and editor with over seven years of experience in a variety of industries. Prior to joining Lightwave and Broadband Technology Report, he was the associate editor of Architectural SSL and LEDs Magazine.