x
Cable/Video

PON at a Crossroads

News on PON (passive optical networking) has been a mixed bag lately. Even as Marconi PLC (Nasdaq/London: MONI) confirms that it's jettisoning its PONs (see Marconi to 'Streamline' Access ), startups such as Iamba Technologies Inc. are set to exhibit PON-capable prototypes at the CeBIT technology trade show next week.

And there's more: Late last month, Ethernet PON vendor Salira Optical Network Systems Inc. announced new funding (see Salira Secures $7 Million More and Salira Puts on Its EPON), while PON market leader Optical Solutions Inc. is reportedly struggling for new backing.

What gives? Has PON, a technique that deploys passive splitters to siphon optical bandwidth among multiple users, hit the wall as an optical access technique? Or is it entering a new phase?

A bit of both, sources say. On the one hand, changes in the market appear to have caused interest in first-generation ATM-based PON products to narrow. It looks like the original hope that RBOCs and incumbents would buy PON in a big way has fizzled.

Instead, the recent pattern of contract wins (see PON: The Dream Is Alive) shows PON is selling largely to independent service providers associated with municipal projects and to real estate developers looking for a fast way to put fiber into new housing developments.

"Capex and unbundling concerns will probably push RBOC plans out six to nine months," says Darryl PONder [Ed: caps ours], CEO of Optical Solutions, a PON market leader. In the meantime, he says, there is a steady flow of demand from small to medium-sized businesses. Ponder also sees a market among so-called multiple system operators (MSOs), who are thought to be seeking to one-up the RBOCs with hot new services that include packet voice, video on demand, and high-speed Internet access (see Cable MSOs Set to Win?).

Ponder says all this has helped Optical Solutions make $10 million in 2001, and he is confident of "more than doubling" that figure this year.

Ponder admits, however, that it hasn't been an easy year. Optical Solutions recently laid off 54 employees, bringing its total employee roster to 152. And the company is reaching the bottom of its $104 million initial funding. Ponder says it's taking longer than he'd hoped to raise money, but he is confident of having an insider round sometime within the next couple of months.

He also thinks the RBOC market will materialize. Bell Canada (NYSE/Toronto: BCE), Qwest Communications International Inc. (NYSE: Q), SBC Communications Inc. (NYSE: SBC), and Verizon Communications Inc. (NYSE: VZ) all have trials underway, he notes. Even if they don't act right away, he expects to see significant RBOC activity by 2004.

Some sources say that's not likely to happen. "PONs work, but you have to put the glass in the ground," says Barry Moon, senior analyst at RHK Inc.. Incumbent carriers, he says, aren't willing to spend a few billion dollars to upgrade a network with fiber in order to extend PON services -- until they can be sure it will make them money.

And that, says Moon, has yet to be proven: To offset the initial expenses, incumbent carriers would have to "convince everyone to spend more for services than they're already paying... You have to ask yourself how long a carrier will stay in business who's charging $800 a month for 300 Mbit/s of bandwidth."

Market figures bear out the thin adoption of PONs as they exist today. In 2001, for instance, PON accounted for a fraction of expenditures on access technologies, compared with Ethernet, DWDM, and Sonet/SDH (see Last Mile Reaches Out ).

But existing PON vendors say there's still a market out there ready to be mined. Quantum Bridge Communications Inc., for instance, sees no slowdown in PON demand among MSOs and regional municipalities, particularly when it comes to business services. "MSO business services are growing at 30 to 40 percent per year," says Jeff Gwynne, cofounder and VP of marketing.

Others say the PON market is still viable, but not as an ATM-based standalone product. Cheaper Ethernet-based PONs (EPONs), they claim, will nudge the incumbents to open their wallets. "Do I think EPON is priced cheaply enough to inspire carriers to invest in PON? Absolutely," says Tom Walsh, VP of sales and marketing at Salira Optical Network Systems Inc., an EPON startup.

ATM-based PON makers are realizing they're being nudged out, and it won't be easy for them to "shapeshift" their way into a new kind of technology, Walsh says. "If I were an ATM PON vendor, I'd be telling people now that there's a need to support both ATM and Ethernet," while in reality, he claims, the old ATM gear will fall by the wayside as more EPON products come to market. Salira's first product is set for release in the second half of this year.

"We're trying not to be smug, but there's a huge interest in optical Ethernet access out there, domestically and internationally," Walsh says.

Other companies, including Iamba Technologies Inc., see PON as just another element of a broad optical access spectrum that's increasingly in demand. Iamba, for instance, says it will offer point-to-point Ethernet, Sonet/SDH, and PON on its multiservice access platform, while supporting ATM and leased lines too. Products won't be ready until the first half of 2003, however.

In the meantime, it looks like the old guard is getting the message. Quantum Bridge recently added Ethernet to its boxes (see Quantum Bridge Casts a Wider (Ether) Net), and the company says its long-term goal is centered on optical access, not just PON. "Eventually, we plan to derive less than half our revenues from PON," says Gwynne. "The rest will come from our multiservice capabilities and WDM."

Gwynne says Quantum Bridge will be announcing a series of partnerships within the next few months geared to "enhancing distribution and driving additional revenues."

— Mary Jander, Senior Editor, Light Reading
http://www.lightreading.com For more information on CeBIT, please visit: www.lightreading.com/cebit

Page 1 / 2   >   >>
rtec935 12/4/2012 | 11:59:50 PM
re: PON at a Crossroads Folks,

I am a new kid in the block. I am learning. Actually, lots of things to learn: ATM, PON, EPON, FSAN GǪ

Here is the first step: Could someone please provide a brief summary of Pros and Cons in regards to FSAN? How it varies with Pt-to-Pt fiber optic architecture in the Access network?

Thanks,
New Kid and wants to learn
edgecore 12/4/2012 | 10:49:19 PM
re: PON at a Crossroads What is the attraction for EPONs vs ATM PONs?

EC
DKP 12/4/2012 | 10:49:18 PM
re: PON at a Crossroads
> What is the attraction for EPONs vs ATM PONs?

EPONs have many advantages;

(1) Ethernet Switch architecture
EPONs will use standard Ethernet chips (switches, etc), and have the advantage of working with Ethernet Switch/Router cores, as well as management software. This saves in cost and complexity.

(2) Standard:
ATM PON never made it to a standard. FSAN spent 7 years and there is no compatibility between equipment. EPON is being standardized as 802.3ah; an Ethernet standard. It will insure compatibility and plug-and-play ONUs.

(3) IP
Ethernet PON is a packet based network, designed to carry IP. Proper system designs allow for voice, data and video delivery.

(4) Point to point subset
What is an EPON with one ONU: It is a point to point optical Ethernet link.

(5) Speed
ATM PON is standardized for 155 Mbps (OC-3). That is too little too late. EPON is 1 Gbps, enough for bundled voice, data and video.

(6) Cost
EPONs can take advantage of an Ethernet Switch designed architecture. EPON transceivers are GBICs, modified for fast on-off. EPONs will take on an Ethernet cost model.

(7) Simplicity
ATM PONs are a swamp of complexity. Picture taking ATM and adding complex algorithms. EPONs have a very simple control algorithm.

(8) End-to-end Ethernet
What is the internet port on your PC: Ethernet. What is the main port on a CO Router: Ethernet. Why the heck would you use ATM to connect these two?

(9) Service Provider Shift
Service providers have shifted to Ethernet. NTT/Usen have announced EPON deployment intentions. Asia PTT are deploying optical Ethernet. US ILECs are starting to evaluate EPON.

And this is why Salira, an EPON company, can get funding in this environment.
topicoptic 12/4/2012 | 10:49:16 PM
re: PON at a Crossroads I was once an Engineer at a PON company for almost two years. Many points made below are incorrect. I don't defend APON or EPON, although my answers might seem APON biased, but I'm just setting the record straight. Both will have their place in the market. ATM centric networks will choose APON and IP centric networks will choose EPON, it is that simple. If you want answers then see below.

T.O.




EPONs have many advantages;

(1) Ethernet Switch architecture
EPONs will use standard Ethernet chips (switches, etc), and have the advantage of working with Ethernet Switch/Router cores, as well as management software. This saves in cost and complexity.

This statement is partially ture, Ethernet components are typically cheaper than ATM components. But it all depends on the level of comlexity designed. Many PON systems are IP or ATM muxs with very little switching. ATM has just as many commercially available components and standards as Ethernet.


(2) Standard:
ATM PON never made it to a standard. FSAN spent 7 years and there is no compatibility between equipment. EPON is being standardized as 802.3ah; an Ethernet standard. It will insure compatibility and plug-and-play ONUs.

FSAN is much further along than EFM for PON systems. FSAN has defined several standards around PON systems architecture, optic capacities, frame format, wavelengths, dynamic bandwidth allocation, etc. EFM has just begun it's battle, of which EPON is one small piece. It will be 24 months before EFM has an EPON standard to the point of FSAN today.



(3) IP
Ethernet PON is a packet based network, designed to carry IP. Proper system designs allow for voice, data and video delivery.

IP is proper system for carrying voice ? Tell that to the RBOCs !

(4) Point to point subset
What is an EPON with one ONU: It is a point to point optical Ethernet link.

I'm not sure of the point being made here. One of PONs benefits is multiple endpoints on the same fiber plant to maximize cost to help drive fiber deeper into the access network. Point to point arguements are irrelevant when talking about PON systems.

(5) Speed
ATM PON is standardized for 155 Mbps (OC-3). That is too little too late. EPON is 1 Gbps, enough for bundled voice, data and video.

APON has two standards for speed 155Mbps and 622Mbps. No it is not OC-3. EPON has no standard speeds because there is no standard. But this discussion is not relevant because the protocol does not dictate the speed of the PON, the optics do. I know APON vendors currently working on 1.25G and 2.5G PON speeds. Also, be careful of advertised speeds. How much payload can AllOptics 1.25G PON carry ? If I have 1.25G lasers but can only carry 500M of payload is that a good system at less than 50% efficiency ?


(6) Cost
EPONs can take advantage of an Ethernet Switch designed architecture. EPON transceivers are GBICs, modified for fast on-off. EPONs will take on an Ethernet cost model.

Partially true. Ethernet can be a cheaper technology but it has nothing to do with fast on-off.


(7) Simplicity
ATM PONs are a swamp of complexity. Picture taking ATM and adding complex algorithms. EPONs have a very simple control algorithm.

This is such a high level statement that I'm not sure of the point being made.


(8) End-to-end Ethernet
What is the internet port on your PC: Ethernet. What is the main port on a CO Router: Ethernet. Why the heck would you use ATM to connect these two?

PON systems demarc at the CPE, not the end user device. Every PON company's ONU has an Ethernet port. The main port on a CO router might be Ethernet and it might be ATM. Or there may not be a router in the CO at all, it may be an ATM switch. It all depends on the Provider.

(9) Service Provider Shift
Service providers have shifted to Ethernet. NTT/Usen have announced EPON deployment intentions. Asia PTT are deploying optical Ethernet. US ILECs are starting to evaluate EPON.

NTT is deploying APON, that is a fact. Optical Ethernet is a nice marketing phrase, what exactly does it mean ? Evaluting and purchasing are two different things. If you truly believe the RBOCS are going to carry voice and broadcast video over IP within the next decade then you have some homework to do. Now, there are several Cable Operators and other IP centric providers that will choose EPON to carry voice.

And this is why Salira, an EPON company, can get funding in this environment.

APON and EPON companies have raised funding just the same. The companies that get noticed and raise money will be the companies that sell product and make money. This has nothing to do with PON. Salira is still 6 months ways from any product. If they have a solid optic engineering team and can get the PON optics to work then there's a chance they'll be OK. The protcol part, ATM or IP, is easy.
rjmcmahon 12/4/2012 | 10:49:13 PM
re: PON at a Crossroads Lifeline POTS is still a requirement, this is not solved by waving ones hands.
____________________

Lifeline POTS should move to health monitoring companies. My mother living alone can't make a 911 call when her heart stops and she didn't notice her TIA. Thank god she was visiting when we noticed her speech slurred.

These can be detected much earlier once their is an always on health monitoring service in her home. That's worth $50 per month and probably more.

dietaryfiber 12/4/2012 | 10:49:13 PM
re: PON at a Crossroads Gentlemen,

I found both of your claims somewhat untrue and perhaps market based. PON in business markets and residential markets is somewhat different. I will add my own spin below:


(1) Ethernet Switch architecture
EPONs will use standard Ethernet chips (switches, etc), and have the advantage of working with Ethernet Switch/Router cores, as well as management software. This saves in cost and complexity.

This is true at some level, however ASICs rule the day in the Telecom world and the cost delta is actually smaller. However, the management piece is untrue depending on who you sell it to. OSMINE is OSMINE is OSMINE. It is not clear to me that an Ethernet based system has even been effectively modeled yet.


(2) Standard:
ATM PON never made it to a standard. FSAN spent 7 years and there is no compatibility between equipment. EPON is being standardized as 802.3ah; an Ethernet standard. It will insure compatibility and plug-and-play ONUs.

There is no interoperability today on any ONUs. FSAN is a much more mature standard, but to date no one has tried to be interoperable. This should let you all understand this is still at the level of sandbox technology.

(3) IP
Ethernet PON is a packet based network, designed to carry IP. Proper system designs allow for voice, data and video delivery.

IP is the primary vehicle for transport, however there are no Class 5 IP switch in the telco world and the Nortel deal at Sprint is based on ATM. So, interworking back into Class 5 is an issue. On the cable side of the business, VoIP would be the proper architecture mixing very well into a DOCSIS network. There is an issue of QoS associated with Video however, and this would need some work.

(4) Point to point subset
What is an EPON with one ONU: It is a point to point optical Ethernet link.

This is strictly meaningless noise. A point to point ATM link is an OC3c or an OC12c. And?

(5) Speed
ATM PON is standardized for 155 Mbps (OC-3). That is too little too late. EPON is 1 Gbps, enough for bundled voice, data and video.

Actually 622 Mb/s APONs are being deployed today in small quantity. 1 Gb/s is certainly not enough for later this decade with HDTV at about 20Mb/s per channel (3 TVs per home and a 32:1 split for economics). Sorry about that, 1.5G is probably what is required here.

(6) Cost
EPONs can take advantage of an Ethernet Switch designed architecture. EPON transceivers are GBICs, modified for fast on-off. EPONs will take on an Ethernet cost model.

The optical work to build off the shelf optics and components is tremendous and new. There is no reason to believe that the optics of one system would not work for the other.


(7) Simplicity
ATM PONs are a swamp of complexity. Picture taking ATM and adding complex algorithms. EPONs have a very simple control algorithm.

Actually, this is simply untrue. It is more complex to get good utilization with variable length frames as this will require additional signaling. The MAC layers of both however can be made in silicon in a pretty straightforward manner and build on other work.

(8) End-to-end Ethernet
What is the internet port on your PC: Ethernet. What is the main port on a CO Router: Ethernet. Why the heck would you use ATM to connect these two?

This is true for a business. However, my home has a TV with a Coax and a phone with an RJ-11 jack. Neither of these devices plug effectively into an RJ-45. Only in the last few years have some homes been wired with CAT5. The resolution of home wiring for any PON or Home Gateway environment is challenging. Wireless may play a role here, but again HDTV will not scale over any home wireless system proposed to date.

(9) Service Provider Shift
Service providers have shifted to Ethernet. NTT/Usen have announced EPON deployment intentions. Asia PTT are deploying optical Ethernet. US ILECs are starting to evaluate EPON.

NTT is also speeding is role out of ATM-based DSL. So, I think that Ethernet is an interesting Access Architecture there is much work to do here. If you have ever met an Outside Plant group at an RBOC, you would find they are uncomfortable without 100% control. Connectionless systems will take some time to overcome that cultural barrier. This is less so in the business environment, and that looks to be the place that PON will start deployment in any case.

And this is why Salira, an EPON company, can get funding in this environment.

Quantum Bridge got some money last fall I believe. I think that people are still enamored by the PON dream.


So, what stops FTTH:

1 - Regulation: If you thought unbundling was bad with DSL, wait to you go PON.
2 - More Regulation: Lifeline POTS is still a requirement, this is not solved by waving ones hands.
3 - Carrier Practices: Access Networking people are not skilled in handling FTTH. It will take time for this to be done.
4 - Cost: This needs to be done in a way that meets the cost/service targets of carriers.

You can see that many of these elements are much less prevalent in the business world than in the residential world. However, the business competition is SONET muxes dropping T-1s/T-3s to customers. Not easily solved, but PON does have some advantages.

dietary fiber
dietaryfiber 12/4/2012 | 10:49:12 PM
re: PON at a Crossroads That is a different service and does not obiviate the need for incumbent local carriers to provide lifeline POTS.

The term lifeline means that it stays in service even when the electrical power is out. This does not mean that you can call when you are incapacitated.

dietary fiber
rjmcmahon 12/4/2012 | 10:49:11 PM
re: PON at a Crossroads Fair enough on the definition of terms. Maybe it should be called powered line instead and then a discusion of where powered lines are required can be debated without the hand waving. Or better yet, let the markets decide by paying extra for powered lines where needed.

My experience has been that incumbants use the lifeline arguement where it hasn't been necessary. They did this to protect their revenue streams, not to service a customer need.
teng100 12/4/2012 | 10:49:11 PM
re: PON at a Crossroads If most of the internet traffic are 40 bytes for
ACK, what is the advantage to use minimum 64 bytes
of the Ethernet packet, and how do you achieve the
efficiency wheh the Ethernet has only 80% of the line bandwidth allocated for the payload after
4b/5b or 8b/10b is taken into the picture.

Neither technology is absolutely better than the other, and whichever works for all the requirements in a targetted service and then it is cheaper will win.
Ted F. 12/4/2012 | 10:49:08 PM
re: PON at a Crossroads This article does nothing to qualify the viability and marketability of PONs!
Having VP marketing guys and CEO's of PON companies telling you how big their market is and why people will buy their product doesn't tell you anything. Analysts are also bad sources for such an answer.

You want to know who's going to buy this product, ask the customers who the PONs are targeting: the RBOCS and MSO's...not one quote from them in this article (bad market or bad reporting by LR???)
Page 1 / 2   >   >>
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE