Operator Views on Next-Gen ROADMs

tmmarvel 12/5/2012 | 4:21:08 PM
re: Operator Views on Next-Gen ROADMs

Given that this will be, with the current network bandwidth usage and revenue/cost trends continuing, an essential problem to solve, it would be interesting to know what the actual difficulties are in getting layers beneath packet switching to adjust to actual application load demands..

While solving the fundamental problem that ROADMs seem to be addressing in itself doesnt seem to be waste of time, given how long these solutions (call them WSON, ASON/ASTN, ODUflex, GMPL(amba)S or whatever) have been in the development without a decisive breakthrough, I guess the worthwhileness of this effort depends on whether the 'theoretical' solution will ever be achievable at economical costs.

Are there any developments in the underlying/surrounding technologies that would now fulfill the promise of packet-aware optics (speaking loosely ie incl. actually digital protocols such as ODUflex)?

Or is the over-provisioning of the layers beneath packet switching just a reality that the carriers and end customer need to live with and pay for?

But as mentioned, looking at the network revenue-cost trajectories, it would seem to get worse from here on without some breakthrough.. 

Pete Baldwin 12/5/2012 | 4:21:09 PM
re: Operator Views on Next-Gen ROADMs

OK, folks.  Here's your chance to vote on whether next-gen ROADMs are a waste of time:

ROADMs: What Do Service Providers Really Want?

neyo 12/5/2012 | 4:21:11 PM
re: Operator Views on Next-Gen ROADMs Theoretically possible. It's been conceptualized for a long time now. Nowhere near maturity though. Nevertheless carriers are still recklessly spending big money on optical switching thinking it will materialize soon.
joferrei 12/5/2012 | 4:21:13 PM
re: Operator Views on Next-Gen ROADMs

How about this:

More packet level routing/switching improves optical bw utilization, but makes the overall $/bw higher--what packet level grooming saves in bw, it costs in making the remaining bw needed more expensive.

Similar, optical bypass of packet level processing lowers the overall $/bw, but also lowers the bw efficiency, requiring more bw in the network--what optical router bypass saves in router ports, it costs via increased optical bw requirements.

Can't the vendors develop a system that increases the packet transport bw efficiency without needing packet layer processing in the carrier network, besides the entry/exit points? Or is that simply not possible, and why?

melao2 12/5/2012 | 4:21:15 PM
re: Operator Views on Next-Gen ROADMs

Hehe. I find funny taking in perspective, what was expect from the optical networking technology, and what actually hapenned.

Back in 99 or 2000, the all optical switch was touted to be the holy grail of the backbone. Now that this type of switch is actually feasable, we see that the optical layer cannot solve it all, nor the routing layer.

I have to agree with brookseven. His view is completely correct. All the carriers and vendors have the same strugle.

paolo.franzoi 12/5/2012 | 4:21:16 PM
re: Operator Views on Next-Gen ROADMs



No, I am not.  I am basically saying - and I have no dog in any of these fights - that there are two conflicts going on here.

Bout #1:  Router Vendors versus Optical Vendors:  They are each trying to suck some percentage of the dollars given to the other guy into their products.

Bout #2:  Transport Orgs in Carriers versus Router Orgs in Carriers:  They are each trying to suck budget from the other guys (as well as prestige).

Both sides of both bouts will put up arguments to bolster their claims.  You will be able to find Tier 1 Carrier folks to back BOTH vendor positions.

I am completely opinionless on how it will actually turn out.  The only thing I can say is that dyanamic optical stuff seems like a stretch for guys looking to eliminate patch panels.




neyo 12/5/2012 | 4:21:17 PM
re: Operator Views on Next-Gen ROADMs Brook,
If you are talking about optimizing the bandwidth using ip optical convergence (or synergy ), over a gmpls uni, we have a long way to go and roi on the investments in roadm and otn will take years. Meanwhile we have developments in form of wson, odu-flex and and what not. Early adopters are going to be frustrated at the recurring investments and no visible benefits.
paolo.franzoi 12/5/2012 | 4:21:17 PM
re: Operator Views on Next-Gen ROADMs


No, if you are Cisco you want to build lots and lots of interfaces because that is good for you.  These are connected to lots and lots of remotes and mesh in all kinds of ways

If you are an optical player you want the routers to have fewer high-speed interfaces and use virtual interfaces to connect.

Nobody is doing dyanmic diddly squat.  What they will end up doing is putting more money in optical or more money in routing.  The Transport guys in the carriers want more money in optical.  The Routing guys in the carriers want more money in routing.

You can find someone in a Tier 1 carrier ready to back almost anything.  What they actually do has mileage that varies.



maxwell.smart 12/5/2012 | 4:21:18 PM
re: Operator Views on Next-Gen ROADMs


Please elaborate on what you mean.  Are you thinking of dynamically setup express wavelengths that bypass the normal routed paths ? You may not be able to keep such wavelengths sufficiently full at all times due to the unpredictability of IP traffic.



rahat.hussain 12/5/2012 | 4:21:19 PM
re: Operator Views on Next-Gen ROADMs

i was wondering where mplape was going .. but now it's clear. digital roadm's are the cure-all.


odo <- who claims digital roadms should be "readms", and odo should sign off as oeo

mplape 12/5/2012 | 4:21:19 PM
re: Operator Views on Next-Gen ROADMs

Yes, we agree that there is a lot of interest on behalf of tier 1s to get rid of patch panels.  What I'm contending, is that all optical, CDC, Next-Gen ROADMs will never be neither economical nor manageable.  And the tier 1s that have deployed large all optical ROADM networks and are hoping that Next-Gen ROADMs will save them are going to have to come to terms with thier decisions or be relegated to lower market shares from leaner, faster operators.  IPoDWDM is neither cost effective nor manageble either.  The only economically viable solution is a digital ROADM.

Sterling Perrin 12/5/2012 | 4:21:21 PM
re: Operator Views on Next-Gen ROADMs

It may be a question of the size of the networks and, therefore, the size of the operator. I have talked with several tier 1 operators recently - beyond ATT and VZ - and there is a lot of interest in the technologies to get away from fiber patching - for a small network it may not be a big deal, but for large networks there's a lot of time and money to be saved by automating the optical layer.


paolo.franzoi 12/5/2012 | 4:21:22 PM
re: Operator Views on Next-Gen ROADMs



The counter argument would be to stop buying mesh routing networks with separate router ports per remote and hang one big mamma port off of a lower layer optical network that shuffles bits around at a much lower cost point.


The real question is the OPEX/CAPEX tradeoff.  Vendors will want a premium for the features.  Carriers will want them for free - okay actually they will want a price cut.

As long as people have not thought about that as a business proposition, somebody will make the stuff - which will lead to another round of unprofitable optical products.



neyo 12/5/2012 | 4:21:23 PM
re: Operator Views on Next-Gen ROADMs

I believe CDC is overhyped.

For any backbone, all operators need are end-to-end wavelengths for their core router connectivity and protecting these wavelengths with diverse routes. 

ROADM is a highly luxurious solution where same result is achieved by simple fiber patching, which can be done at the time of implementation. There is no utility for ROADM beyond that.

There is no place for OTN switching/grooming in the network where we already router mesh connectivity at 10G, 40G and 100G per wavelength.

Operators who have a lot of CAPEX to spend will still buy into these new technologies which will make them look cool. 


DCITDave 12/5/2012 | 4:21:23 PM
re: Operator Views on Next-Gen ROADMs

Are carriers asking for something that's just not possible? Or something that's not feasible if you want to run a business?

mplape 12/5/2012 | 4:21:24 PM
re: Operator Views on Next-Gen ROADMs

Next-Gen ROADMs, or Colorless - Directionless - Contentionless (CDC) ROADMs are a bridge too far.  Even if vendors could successfully develop a CDC ROADM, They cannot thoroughly implement a control plane for CDC analog optics elements.

Sign In