GPON is a great technology, GPON stands for gigabit passive optical network, and it is the predominant technology in place that is delivering fiber last mile broadband. The GPON standard was first ratified in 2003, but like most new technologies, it took a few years to hit the market.
GPON quickly became popular because it allowed the provisioning of a gigabit service to customers. A GPON link delivers 2.4 gigabits downstream and 1.2 gigabits upstream to serve up to 64 customers, although most networks I’ve seen don’t deliver to more than 32 customers.
There is still some disagreement among ISPs about the best last-mile fiber technology, and some ISPs still favor active Ethernet networks. The biggest long-term advantage of GPON is that the technology serves more customers than active Ethernet, and most of the R&D for last-mile fiber over the past decade has gone to PON technology.
There are a few interesting benefits of GPON versus active Ethernet. One of the most important is the ability to serve multiple customers on a single feeder fiber. PON has one laser at a hub talking to 32 or more customers. This means a lot less fiber is needed in the network. The other advantage of PON that ISPs like is that there are no active electronics in the network – electronics are only at hubs and at the customer. That’s a lot fewer components to go bad and a less repairs to make in the field.
We’re now seeing most new fiber designs using XGS-PON. This technology increases bandwidth and delivers a symmetrical 10-gigabit path to a neighborhood (for purists, it’s actually 9.953 gigabits). The technology can serve up to 256 customers on a fiber, although most ISPs will serve fewer than that.
The biggest advantage of XGS-PON is that the electronics vendors have all gotten smarter, and XGS-PON is being designed as an overlay onto GPON networks. An ISP can slip an XGS_PON card into an existing GPON chassis and instantly provision customers with faster broadband. The faster speeds just require an upgraded ONT – the electronics at the customer location.
The vendors did this because they took a lot of grief from the industry when they converted from the earlier BPON or APON to GPON. The GPON electronics were incompatible with older PON, and it required a forklift upgrade, meaning a replacement of all electronics from the core to the customer for the upgrade. I helped a few clients through the BPON to GPON upgrade, and it was a nightmare, with staff working late nights since neighborhood networks had to be taken out of service one at a time to make the upgrade.
The other interesting aspect of XGS-PON is that the technology is also forward-looking. The vendors are already field-testing 25-gigabit cards and are working on 40-gigabit cards in the lab. A fiber network provisioned with XGS-PON has an unbelievable capacity, and with new cards added is going to make networks ready for the big bandwidth needs of the future. Any talk of having online virtual reality and telepresence can’t happen until ISPs can provision multi-gigabit connections to multiple homes in a neighborhood – something that would stress even a 10-gigabit XGS-PON connection.
XGS-PON is going to quickly open up a new level of speed competition. I have one new ISP client using XGS-PON that has three broadband products with download speeds of 1, 2, and 5 gigabits, all with an upload speed of 1 gigabit. The cable companies publicly say they are not worried about fiber competition, but they are a long way away from competing with those kinds of speeds.
I’m sure GPON will be around for years to come. But as happens with all technology upgrades, there will probably come a day when the vendors stop supporting old GPON cards and ONTs. The good news for ISPs is that I have a lot of clients that have GPON connections that have worked for over a decade without a hiccup, and there is no rush to replace something that is working great.
Great article. The largest caveat to all this is the transport technology is now way ahead of the curve. I’m still hopeful the various funding methods and grants will aid in bringing fiber to larger swaths of the nation. Also, at this point most networks and protocols have little to no need for such large data pipes. My fear is little demand will lead to a lag in innovation. Exciting times in the communication industry. Thank you for your articles. I’m an avid reader.
The time to Say Farewell to GPON will be there sooner than your expectation. The maturing of 10G pon and competition will drive the carriers to do that. The life cycle of Gpon is no where with a decent upload speed as a whole in the industry because of the dominant position of cables. So, no need for more advanced broadband is the delusion.
Doug, what about NG-PON2?? Is Verizon the only ISP using that platform? Also, are newer ONTs not supporting RFoG video and would that hinder upgrades on legacy systems?
Thanks for this article. Few comments on this
– The fiber itself is not the main cost when deploying a fiber network. +70% goes to digging (at least in Western Europe where no fiber is built on poles) and maximum 10% goes to passive material. I know about p2p networks operated with GPON – splitters are at the POP site
– I assume that any PON technologiy has an other important advantage: requires less power
– However, the vendors managed to keep their connection from CO to ONT proprietory which means vendor locking for an ISP. I guess the TCO over life-time exceeds finally active ethernet – any studies available on this?
It is great to see the offers beyond 1G are coming more and more on the market. Is there and I guess online VR gaming and holograms will be coming soon. Do you have any information about required bandwidth?