GPON Gets a 10G Look

LAS VEGAS -- NXTcomm 2008 -- Ericsson AB (Nasdaq: ERIC) is showing off a proof-of-concept demo of 10 Gbit/s GPON, hoping to boost the access presence that it garnered through the acquisition of Entrisphere. (See Ericsson Demos 10-Gig GPON and Ericsson Spells Defense G-P-O-N.)

Ericsson says this is all about the carriers that are hoping to get more out of big fiber-to-the-premises (FTTP) bets.

“It provides a roadmap for GPON as a sustaining life cycle, said Arun Bhikshesvaran, CTO of Ericsson North America. “I think as we look to a combination of new Internet services and HD channels, these applications will gobble up the capacity.”

Carriers at NXTcomm have demonstrated they can deliver a whole lot of bandwidth on older technologies. Verizon Communications Inc. (NYSE: VZ) announced that it will soon start offering 50 Mbit/s to all of its FiOS customers. (See Verizon Takes 50-Mbit/s FiOS to the Masses.) Most of that bandwidth will be delivered over 622 Mbit/s BPONs.

A beefier GPON, though, could deliver bigger numbers. “Imagine getting 100 Mbit/s to every desktop in an office building. The solutions for that are not currently there,” said Bhikshesvaran.

Ericsson says it increased the capacity of its GPON technology much in the same way carriers increase capacity through copper pair bonding. In this case, it aggregates wavelengths to increase the capacity. “The challenge is to make sure you can retain the photonic performance when you split it across multiple wavelengths,” said Bhikshesvaran.

Standards for 10-Gbit/s GPON aren't ready yet, so the technology -- including Ericsson's demo -- is a ways off from commercialization. The Full Service Access Network (FSAN) organization has a working group established, and it could be in a position to send a recommended 10-Gbit/s standard to the International Telecommunication Union (ITU) in 2010, says Geoff Burke, director of marketing for Calix Inc. (NYSE: CALX)

Meanwhile, the Ethernet PON (EPON) camp continues its work on a 10-Gbit/s standard, as defined by the Institute of Electrical and Electronics Engineers Inc. (IEEE) 802.3av task force. Talk of a 10-Gbit/s EPON has been around for a couple of years, and the standard appears to be on track for ratification in late 2009. (See EPON Evangelists Talk 10-Gig and EPON Goes to 10 Gbit/s.)

— Raymond McConville, Reporter, and Craig Matsumoto, Senior Editor, Light Reading

Page 1 / 2   >   >>
Keebler 12/5/2012 | 3:38:03 PM
re: GPON Gets a 10G Look Blech. Message board screws up html.

Go to fttxnews.blogspot.com and look at Friday, June 20th.
Keebler 12/5/2012 | 3:38:03 PM
re: GPON Gets a 10G Look Interestingly, there is a comment here at the FTTX News blog from someone claiming to be an Ericsson engineer who states that this was truly 10G PON (not DWDM). Perhaps some intrepid LightReading reporter could double-check?
opticalwatcher 12/5/2012 | 3:38:03 PM
re: GPON Gets a 10G Look It's pretty small; instead of a link, just copy/paste it here:


I am one of the makers of the system and can confirm that this system is not hybrid; it runs 10G scrambled NRZ on a single lambda using the G.984.3 GEM/GTC specs at four times the speed of a regular GPON.

Elmar Trojer, Ericsson Research Sweden
[email protected]

6/22/2008 4:51 AM

nodak 12/5/2012 | 3:38:06 PM
re: GPON Gets a 10G Look What would be slick on DWDM system is for the services like TV, you have a tunable RX at the home and some computer smarts to know what wavelength has which channels so you can avoid the duplications. You could then use other dedicated wavelengths for fixed connections, or, if you want to add even more smarts to the home eq, have the home eq look for the lambda with the most bandwidth available.
ponnnn 12/5/2012 | 3:38:10 PM
re: GPON Gets a 10G Look Of course 8 customers per 2.5G (WDM PON) will mean if 3 people on the PON demand 1G of downstream traffic you are in trouble if they all happen to be on the same lambda. On a 10G PON you are fine. Same goes for multi-cast services, if 4 people all want to watch the same show, but they are different lambdas, the OLT needs to make 4 copies and burn the bandwidth on all 4 lambdas.

There is also the additional operational complexity of making sure the ONTs have the proper optics to match the appropriate lambda.

Of course if the price is right, it could all be OK. Or perhaps we will see WDM PON for business applications where bandwidth demand is more constant in theory and the operation complexities are outweighed by the higher revenues?

All in all it will be interesting to see if the PON folks move to 10Gbps (~300Mbps / ONT), how will the copper folks (cable and AT&T) compete with that?
paolo.franzoi 12/5/2012 | 3:38:10 PM
re: GPON Gets a 10G Look
Agreed with both of those.

Duh! 12/5/2012 | 3:38:11 PM
re: GPON Gets a 10G Look Nothing is 'easy'. But no fundamental breakthroughs are needed in OLT receiver design to make 2.4G work with 80 ns guard+preamble+delimiter. 10G is a completely different beast.
paolo.franzoi 12/5/2012 | 3:38:11 PM
re: GPON Gets a 10G Look

I am not sure that a 2.5G upstream is an "easy" upgrade - unless we greatly relax the acquisition time between grants.

Duh! 12/5/2012 | 3:38:12 PM
re: GPON Gets a 10G Look Slick press release and interview... but what aren't they saying? It doesn't take much to get from a working 2.4/1.2G OLT and ONT to 10G/1.2G, and not much more than that to get to 10G/2.4G. Siemens (now N-S) announced they had one of those working in the lab last May, done under an EU project.

Now, if that were 10G upstream, it *would* be an accomplishment.

opticalwatcher 12/5/2012 | 3:38:14 PM
re: GPON Gets a 10G Look Isn't XAUI 4x3.125GHz considered 10Gig? You can argue about terminology, but I think it is much more interesting and useful to talk about technology, market, and costs.

Verizon is busy wiring up the US with BPON and GPON. It looks like they are about to wire up NYC. rjmcmahon is fond of telling us how many years (decades?) it is going to take Verizon to recoup these costs. So just when do you think Verizon is going to rip out all these customer boxes and CO boxes and replace them with much more expensive 'single-wavelength 10G' PON systems?

...especially when there are much less expensive alternatives to get extra bandwidth to the customers that don't require replacing any boxes at all.

I'm suggesting that the IEEE and FSAN commitees are wasting their time with 'single-wavelength 10G' standards work. But like I keep saying. I'm not an expert in these technologies. I'd be interested in hearing alternative viewpoints.
Page 1 / 2   >   >>
Sign In