WDM-PON Faces 10G Challenge

With GPON just starting deployments, vendors are already looking forward to -- and arguing about -- what will eventually be the endgame when it comes to PON.

For a while it seemed that the consensus winner would be WDM-PON, but lately, 10 Gbit/s GPON has been picking up steam, with some feeling it could win outright over WDM-PON years down the road. (See GPON Gets a 10G Look and Who Makes What: GPON Equipment.)

"10 Gbit/s GPON is probably going to be the technology winner because it'll be the right bandwidth at the right costs," says Marcus Weldon, CTO of fixed access activities at Alcatel-Lucent (NYSE: ALU).

That's because bringing down the prohibitive costs of WDM-PON won't be easy. "We see that WDM-PON is currently riding in a two to threefold cost premium to GPON," Weldon says. "To come from two to three times cost down to something that approaches equivalence is a long road. It's not just tweaking components, it's coming up with new components."

Moreover, 10-Gbit/s GPON is a simpler upgrade, since it almost certainly will work on top of existing GPON deployments. "With WDM-PON, you need to rip out the old ONTs, which is a fairly significant overhaul," says Weldon.

Proponents of WDM-PON argue that ripping out and replacing GPON is not the intention of WDM-PON in the first place.

"The belief we have is that the solutions are complimentary," says Giovanni Manto, leader of Ethernet fiber access solutions at Nortel Networks Ltd. , which has a stake in WDM-PON via Novera Optics Inc. (See Nortel JV Buys WDM-PON Specialist.)

Manto notes that GPON has not been deployed in any significant numbers yet and that some service providers still have not made up their minds yet on a fiber access technology.

The WDM-PON camp thinks scalability will help the technology win in the long run. "Shared networks have always been shown to break apart," Manto says. "In addition, with WDM-PON you have the security benefits, because the traffic is not mixed." This is especially important for business and enterprise customers who prefer a point-to-point connection rather than a shared network.

"What we're proposing are point-to-point networks that are scalable," says Manto on WDM-PON. "In my opinion, it also brings a whole new level of service. You have one infrastructure that is simple for enterprise, business, and residential applications."

In the end, according to other vendors, it'll all come down to cost. Tellabs Inc. (Nasdaq: TLAB; Frankfurt: BTLA), as part of the Sardana project, is exploring ways to lower the costs in developing a next generation PON product whether it be WDM-PON or 10 Gbit/s GPON. (See Tellabs Lays Out WDM-PON Plan.)

"As we do the research, if we determine the costs of 10 Gbit/s GPON are cost effective enough, we may decide to pause and operationalize around that," says Tom Ruvarac, director of product management for access products at Tellabs.

One direction Tellabs is studying involves tweaking the types of components placed in the ONTs. For example, it is exploring the use of non-active optical amplifiers that essentially act as mirrors and reflect the optical signal back up to the central office, as opposed to the current laser-based options that are more expensive.

"It should be cheaper than a laser," says Ruvarac. "If I can reduce one of the big cost components inside the ONTs, I can reduce the whole cost of the network."

— Raymond McConville, Reporter, Light Reading

OldPOTS 12/5/2012 | 3:34:11 PM
re: WDM-PON Faces 10G Challenge As an Enterprise Operator I liked the discussion in this article. But it is not just the cost of the ONT that must be reduced, but the labor costs. You can only afford to upgrade once, if at all. WDM-PON will probably win out but only gradually and probably in a decade. But Enterprises will be first as usual for high capacity.


PS There is another related popular thread on this issue.
opticalwatcher 12/5/2012 | 3:34:10 PM
re: WDM-PON Faces 10G Challenge There are so many flavors of WDM that if you have a 10Gig optics PON system to sell it would be pretty easy to make a Powerpoint slide showing your system to be cheaper than WDM.

I'm not an expert, but CWDM seems like an incremental cost increase over non-wave multiplexed PON. And you don't necessarily have to change any electronics if you go to CWDM because both sides would run at the same speed using the same protocol. This is the exact opposite of the argument that AlcaLu uses. How can you use the same electronics if you go to 10Gig speeds?

And considering how long it has taken COs to go from BPON to GPON, and the lack of competition, I don't see any near-term reason for Verizon to change anything, unless, as OP says, there is a need (and the money) in Enterprise networks.
rs50terra 12/5/2012 | 3:34:09 PM
re: WDM-PON Faces 10G Challenge There is another solution to this dilemma: P2P Ethernet. This solution is more cost effective, does not necessitate any complex technical developments and obviates the need for a lengthy process of developing new standards. With each customer physically separated, different rates (100M, 1G and even 10G) can be allocated to each customer according to the service required and subscription paid. No complex bandwidth allocation provisioning is necessary.
Let's remember that one of the strongest arguments of DSL against cable is the shared media characteristic of cable. Why introduce the same issue in the new optical, forward looking network?
Sign In