Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [802.3_GEPOF] FTTH Council market study



Steve-

Thanks for your input.

To the POF group I would say that Steve's point-of-view represents the dominant point-of-view that you will be facing in your tutorial.
You are not facing an audience that does not already have:
- A significant body of satisfactory experience with UTP
- Entrenched established interests in UTP based systems
(Training, tooling, supply chain, cost optimization techniques, etc., etc.)
You are chasing a moving target. 
As I type this your colleagues are formulating a project for speeds higher than 1G over 100 meters of the installed base of cabling.

You need to focus your pitch on the areas that UTP has has enough problems with to justify the conversion into and back out of the optical domain
or where UTP is not a contender at all.
Traditionally, the only areas that have justified that (for glass media) have been electrical isolation and long reach.
(I'm not claiming those are your areas, I'm just citing the obvious example)

Geoff Thompson

On Sep 23, 2014, at 10:04 AMPDT, Steven B. Carlson <scarlson@xxxxxxxxxxxxx> wrote:

All,
 
An interesting thread. I must say I agree with the positions taken by Bob, Geoff and Marek.
 
I’d like to clarify the use on Type NM (non-metallic) wiring in single-family residential construction. Type NM, usually called “Romex,” is a 2 (hot and neutral) or 3 (hot neutral, and earth ground) conductor solid conductor cable with a non-metallic sheath. Romex is a trademark of Southwire Corp. and was developed in 1922 by Rome Wire and Cable (now part of Southwire.) It was developed to replace Type MC, e.g. BX cable, which had a metallic sheath. BX cable is still in use in commercial construction as Type AC (armored cable.)
 
I agree with Bob that Type MN has been used in North American residential construction (USA and Canada) for at least 60 years. It is not placed in conduit within interior walls, and may be used with cinder block or cement block outer walls. Type NM may be used in construction up to 3 stories, which covers apartment complexes, condominiums, etc. Above 3 stories, commercial construction rules must be used.
 
Even if conduit was used in North American residential construction, the National Electric Code would have to be amended to allow POF in the same conduit, as it would have to meet the same fire and smoke production requirements as Type NM.  In addition, there are city, county and state codes that may affect the use of POF, or any other media. One size does not fit all. I have some experience in this area as I have worked with NEC Code Panel 15 off and on for over 30 years.
 
It’s worth noting that electrical codes in North America focus on fire prevention, while those in Europe focus on electrocution hazards.
 
It should be possible to get a rough idea of the installation of structured wiring plants over the last 15 years, but it would be a lot of work and a number of different entities would need to consulted. I would suggest asking our cabling experts if they have such data.  I think Marek is spot on with his comment---it all  depends on where you are and who you ask.  Structured wiring is typically used in higher-end homes, where it also carries telephone (POTS and IP), security, HVAC, and lighting control, and Ethernet traffic. Most of the applications are non-network (or at least Ethernet) and rely on the fact that this is copper wire that can carry power and control signals. It should be noted that much of this equipment is moving to wireless to support the much larger retrofit market. Home automation and security has moved largely to wireless, and young people today, who have had a cell phone since they were 8 years old, don’t own a conventional home phone. For those of us who want a structured wiring plant, they can easily be added to older homes as it does not involve electricians, permits, or electrical inspectors. I added it to my 1978 home 10 years ago, and it took 2 installers 4 hours to install, test, and sign-off on the job. Carlos and I might take 10 minutes to do a 110 punch-down termination, but these professionals could do one in a little over a minute!
 
You will fight a losing battle if the focus is on the idea that POF will replace all other media everywhere, in all applications. Instead, focus on those applications where there is a real (not imagined) advantage, which would be dwellings with thick stone walls and exposed electrical conduit, or MTU where existing conduit may be used.
 
I’d also be curious if Eugene can supply some hard data on EMC issues---I have been building large Ethernet networks for theme parks, convention centers, Las Vegas showrooms, movie studios, etc. using UTP (and MMF for multi-kilometer links) for 25 years, and have not seen any problems with UTP. I would expect that EMC that severe would cripple any electronic equipment in the vicinity.
 
Best regards,
 
Steve
 
 
Steven B. Carlson
Chair, IEEE P802.3bp 1000BASE-T1 PHY Task Force
http://www.ieee802.org/3/bp/index.html
Executive Secretary, IEEE 802.3 Working Group
http://grouper.ieee.org/groups/802/3/index.html
President
High Speed Design, Inc.
 
 
From: Marek Hajduczenia [mailto:marek.hajduczenia@xxxxxxxxx] 
Sent: Tuesday, September 23, 2014 6:47 AM
To: STDS-802-3-GEPOF@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_GEPOF] FTTH Council market study
 
Carlos,
 
Spelling out these target application areas in the project documentation would be helpful, and that would address my concern about the target market for this project. This would clearly identify the areas of opportunity for this project.
 
Thank you
 
Marek
 
From: Carlos Pardo [mailto:carlospardo@xxxxxxxxx] 
Sent: Tuesday, September 23, 2014 9:16 AM
To: Marek Hajduczenia; STDS-802-3-GEPOF@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_GEPOF] FTTH Council market study
 

Dear all,

I do agree with Marek and Geoff in the sense that we should present market numbers which expands the current Ethernet numbers. CAT-5e/6 is the main media, and we should avoid giving the message of POF will cannibalize the copper market. Which, I think, even if we want, we will not be able to do. CAT-5e/6 is the regular media for all the wired Ethernet devices.

When the CFI was presented, the market opportunity shown, was the stone/brick houses or multi-dwelling apartments, where the mains are distributed through plastic conduits. This is the case of many homes around Europe. And this is why Europe will lead this market. Of course there might be some opportunities in US, but they will not be driven by how homes are build in average in the US. Eugen is trying to identify these opportunities, ( EMC, Galvanic, etc). But, again, should be new market opportunities, not the old copper market now to be shared with POF. This will never happen. Opto-electronics is more expensive than copper link magnetics, and a POF link will be more expensive than a copper one. Is naive, in my opinion, to think that POF will take over copper. Again, even if we would like that ( which I don't), it will never happen.

We should concentrate in the solution we want to provide. Now in Europe the FTTH deployment is arriving to most of the big cities. Customers receives 100 Mbps or more, and they can only use it in the same room where the WiFi router is located. Going beyond one or two walls, the coverage is very small, or even no connection at all.

POF can be installed in all this houses in parallel with the FTTH deployment. It will be done reusing the mains plastic conduits, to provide connectivity in every room. These are the "brown field" houses. As an example, Telefonica use to install a directional, point to point propietary WiFi solution between the router and the STB, when the STB is placed in a different room than the router. This is a expensive solution and it does not work in all the cases. Also the access speed provided by Telefonica is growing, and high speed connectivity is required in all the rooms.

For the new ones ( "green field" ) there might be also opportunities, but in this case, the economics should come from the money saved of avoiding the installation of two plastic conduits ( one for the mains and another for data ), vs one (POF + mains), as well as,  may be, installation time, vs the cost of active wall plugs.
I am not sure about it, and we should be very certain about these numbers before presenting it as a real market.

Regarding the PoE, at least for the European market, I think is not a problem. We are targeting mains conduits, so power will be there for the wall plugs. It is also true, that these wall plugs should provide PoE to the devices connected to it.

Best Regards

Carlos Pardo

El 23/09/14 13:34, Marek Hajduczenia escribió:
Eugene,
 
I would very much like to see the exact conditions and the reason for these quoted houses to RF interference that prevents the use of twisted pair. Such levels of RF interference are not high enough to prevent normal Ethernet operation even in the vicinity of the base station, let alone inside of a residential building. I am curious what might be the root cause there. I do not believe in jumping to solutions without knowing specifically if the existing ones would not work.
 
I would also note that there is *no* actual data on how many homes are installed with dedicated structured wiring today in any market (unless I missed it altogether), and my personal estimates in the area where I live range from 100% (dedicated fiber communities, structured wiring is part of the base package) to ~15% (structured wiring is offered as an optional package to a home buyer). There are plenty of retrofits being done in houses, especially when walls go down and if in the meantime people discover that WiFi does not cut it in all situations (building structure, distance from AP, building materials, interference from neighbors, etc.). Given the spread I observe in the area I live, one can easy claim twisted pair in the house is very common or very rare, depending on where you live and what data sample you take for reference. I believe we need a data set to look at before we make a leap of faith and assume there very little twisted pair in the house (or vice versa).
 
It would be also nice to get opinions from home developers, people who are said to have so many problems with twisted pair today. I spoke with a few of them in the area where I live, and perhaps it is a Florida thing, but none of them were concerned about twisted pair, their ability to offer it in the houses, or problems with quality of resulting structured wiring. What they did dislike, though, is the active sockets on the walls, we have already discussed. I understand that it is the *first* generation of products, but if my understanding is correct, to make the wall socket all passive, we now would need either an external converter from POF to twisted pair for consumer electronics (that is what comes on TV, PCs, laptops, etc.), or require that each new piece of consumer electronics be equipped with a POF port. I do not see that happening: consumers want fewer boxes in their houses (hence the active wall socket), and manufacturers of consumer electronics already have high speed interfaces that work (HDMI, DisplayPort, USB3.1, and others). This is a very competitive environment that crossed 10Gb/s barrier some time ago, and I do not believe that entering this market with an interface that does only 1Gb/s would be compelling to anybody.
 
Don’t get me wrong, Eugene. I think the use of POF might make sense when you retrofit the house and have absolutely no access to the inside of the walls, and have to run data cabling in the same conduit as the power lines. However, I would not jump to conclusion that POF will take over the world, and become a standard for home networking. I believe it has its application, and the project just needs to qualify the market segment it is going after, and not go for everything (and nothing).
 
Regards
 
Marek
 
From: Dai, Eugene (CCI-Atlanta) [mailto:Eugene.Dai@xxxxxxx] 
Sent: Tuesday, September 23, 2014 2:17 AM
To: STDS-802-3-GEPOF@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_GEPOF] FTTH Council market study
 
 
Geoff: it's better to got pain now..., see my answer inline...

From: Geoff Thompson <thompson@xxxxxxxx>
Sent: Tuesday, September 23, 2014 12:54 AM
To: Dai, Eugene (CCI-Atlanta)
Cc: STDS-802-3-GEPOF@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_GEPOF] FTTH Council market study
 
Eugene-
 
Just to be a pain in the A--
(Because it is better to get the darts now than at the Plenary
 
See below
 
Geoff Thompson
 
On Sep 22, 2014, at 8:52 PMPDT, Dai, Eugene (CCI-Atlanta) <Eugene.Dai@xxxxxxx> wrote:

 


The market of POF for home networking in my opinion is referred as a global market, lead by Europe and other regions at this moment, not North American. The requirements come from several areas:
  • Bandwidth requirement for home networking in the future: Advance in FTTH technology enables gigabit services to the home, old home network physical medium has become or will become the bottleneck
GOT:

I don't understand your point here.  Are you trying to argue that the installed base of Cat5 in the residence will have to be replaced?  Are you trying to argue that POF will win over UTP on a level green field?  If that is your tactic then you are intentionally trying to annoy the incumbent forces in 802.3.  That seems like a bad idea if the goal is to get 75% them to vote for your project.
ED: No, POF will not replace anything that already exist. Just like the very limited CAT 5 in residence does not replace phone line or coax.

  • EMI and RF noise environment: POF for DOCSIS HFC home networking is already used by an US MSO to solve the noise problem otherwise difficult to solve.
GOT:

Again, I don't understand your point.  Are you trying to assert that POF's place in the market is as the final link in an access network or is its place for backbone wiring within a building from the demarc to the outlets?  If you don't have real focus and know what you are shooting for, you will be ripped to shreds.

ED: POF will not be in access network. This is a statement of fact that POF already used today in US for some home networking case where noise impairment  s hard to correct.

  • POF has potential to provide bandwidth grows for home networking in the future.
GOT:
So does everything else.  UTP and glass fiber (multi-mode and single mode) have already been proving that for years.  A simple assertion that this is the case for POF will not be taken as a given by this group.  Further, the level of investment to improve modulation for POF is (at least at the current point) is significantly smaller than for the incumbent technologies.

ED:When easy installation, termination of fiber is the key weight, POF is the choice for home networking over SFM or MMF

 

  • How devices been built is a product implementation issue, not related to IEEE standard.  The industry can start with active GEPOF wall jack; gradually move to passive wall jack with media convert build in the consumer devices. The only thing we have to prove is that the product could be build with relatively low cost.
GOT:
The issue here for your INITIAL Broad Market Potential is:
How will the initial GEOPOF wall jack have BMP if it needs a wall wart? Rather than how that will be achieved after you convince a significant portion of the Ethernet appliance manufacturers to build their consumer devices with PoE back feed (this will add on a cost probably larger than the cost of the Ethernet TP interface itself).

ED:Active wall jack POF converter product already exist. How product evolve to the next stage, such as integrated GEPON interface in consumer device is market driven, not standard driven.

 

  • The entire POF for home networking has nothing to do with how the home is been built or with what kind of materials. It could be built with paper, doesn't matter. It about the ability to meet the bandwidth growth, easy termination and installation, safety, EMI, RF noises…
GOT:

You are welcome to think that but your audience strongly believes that POF is a late arrival to a battle that (a) has already been won by UTP and (b) that there is a large installed base of Cat5 and (c) has already won the default interface for Ethernet equipment in the world.

ED: I don't there is a large installed CAT 5 in home.

  • To IEEE, GEPOF means Ethernet runs on a new medium
GOT:

This is only interesting to 802.3 if it expands the Ethernet market more than it further fractures the market into segments:

  • To home builders it means more choices for wires
GOT:

And why is this good?  This gives a homebuilder choice in an area where he doesn't want it.  Homebuilders need to devote their effort to making the choices in the area that their customers have opinions (floor plans, bathroom finishes) For infrastructure issues, builders wants life to be simple and things to be a given.
ED: Again if bandwidth requires fiber in the home, builders have to choose GOF or POF, and POF is much easier to handle.

  • To service provides it means a new technology in their tool kit
GOT:

And one more thing to train technicians for and one more set of tools to carry in their tool kit.
ED: We do this every day.

  • To consumers it means a new interface, just like HDMI, or optical interface for DVD player...
GOT:
And each new interface is a major pain.  Consider the screaming that went on when Apple changed the interface on the iPhone.  The major advantage to TP Ethernet is that the same connection has worked for 25 years and the packet format hasn't changed in 34 years (see also USB, VGA, RJ-11, etc.)

ED: Then don't ever use HDMI, go back to coax for HDTV TV, 4k TV.... Hope it will work.
 
Good luck.
 

 

Eugene

From: ROBERT GROW <bobgrow@xxxxxxx>
Sent: Monday, September 22, 2014 10:24 PM
To: STDS-802-3-GEPOF@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_GEPOF] FTTH Council market study
 
I’m not personally aware of statistics on construction techniques.  I strongly believe that most all single family homes built in the last 60+ years have used wood stud walls in most parts of the US (I believe Canada is similar but am less confident in that assertion).  As Mr. Thompson noted, Florida may be the exception. where block construction is much more common (Florida has a reasonably strong justification for construction resistant to hurricane force winds).  
 
Condominium and low rise apartment buildings also are likely to be wood stud construction.  High rise apartments are much more likely to use conduit.  Unfortunately, I don’t personally have data on even this distribution of the US population (single family home verses multi-tenant, though a web search might provide some data).
 
My main point on the second bullet was that what you ask about is a vendor product issue to the 802.3 working group, not a standards project issue.  A GEPOF amendment to 802.3 would not even mention the types of devices that might exist at an electrical wall outlet.  Informing members of 802.3 of the possibilities for such products helps them understand how use of the electrical conduit (where it exists) is very practical for building a full coverage home network.  I personally don’t think whether the POF termination at an outlet is only a Wi-Fi device, or if it is a converter to 10/100/1000BASE-T, or if it is both Wi-Fi and BASE-T is important to project approval.  The important point is how POF solves a real and serious installation problem for building an Ethernet home backbone in various geographic regions (e.g., Europe).
 
— Bob
 
 
On Sep 22, 2014, at 3:40 PM, Vikas Sarawat <v.sarawat@xxxxxxxxxxxxx> wrote:

 

Thanks Bob. 
 
I find your email on bullet#1 very informative. Do you know of any resource where I can find some statistics on how many homes in the US have Romex vs conduit vs something else?
Do we have similar information for homes in Europe?
 
On bullet number two, my question is:
Since we are solving a real word problem, for POF enabled homes what will the outlet look like? Is this active? Do we expect to have media converter on each outlet? Do we expect consumer devices to have media converter or support POF natively?
 
Thanks,
Vikas
 
From: ROBERT GROW <bobgrow@xxxxxxx>
Date: Monday, September 22, 2014 at 10:44 AM
To: Vikas Sarawat <v.sarawat@xxxxxxxxxxxxx>
Cc: "STDS-802-3-GEPOF@xxxxxxxxxxxxxxxxx" <STDS-802-3-GEPOF@xxxxxxxxxxxxxxxxx>
Subject: Re: [802.3_GEPOF] FTTH Council market study
 
1.  The first question is certainly relevant to market potential.  But construction techniques make POF and Cat-5 equivalent for most homes.
 
Most current US home construction does not use conduit.  The houses I’ve owned over the decades in California, Connecticut, Pennsylvania, Ohio, and Utah didn’t have electrical conduit.  Electrical wiring for all was with a product called romex.  Widely used since the mid 20th century.  Walls, both interior and often exterior, are wood stud.  Exterior walls will be insulated, but interior walls rarely are insulated.   
 
Romex is typically stapled to wood studs (code specifies how close to outlet, junction and switch boxes the electrical cable needs to be stapled).  For horizontal runs, a hole is drilled in the stud and the romex is threaded through.  There is no way to use the routing of romex in walls as an aid for network cabling installation after construction is finished.  
 
Multi-dwelling units are much more varied in construction techniques.  Some apartments, condos and town homes have cinder block, brick or other solid wall separation between units, but they might also be build with wood studs and insulation between units.  Depending on building height, floors might be wood or poured concrete.  For concrete, conduit would be common.  (Many California homes are built on a poured concrete slab, and conduit may be used there.)
 
Business buildout of large spaces is often done with metal studs.  An alternate electrical cable is often used here, a flexible metal jacketed electrical cable.  This isn’t a conduit and therefore when used in business or home construction doesn’t provide any aid for network cable installation.
 
Because of the size of POF, it has a similar installation for the US.  Rather than going into walls, sometimes it is easier when installing a home theatre system to simple gouge into the wall (usually wall board, install the cable and then seal in the cable.  This isn’t an option with standard Cat 5 because of cable size.  Similar for under carpet installation, though flat cabling is an option.  I believe POF is a viable option for the US, but it won’t have as much conduit to utilize as it will in other markets.
 
2.  Not sure how your second point relates to an 802.3 standard.  We would not specify outlet box Wi-Fi APs, nor any other elements of the home network architecture.  The only relevance is in the number of POF ports involved in a home network.
 
3.  Future POF is in the presentation, and will be an item of interest.
 
—Bob
 
On Sep 22, 2014, at 7:29 AM, Vikas Sarawat <v.sarawat@xxxxxxxxxxxxx> wrote:

 

These are good points.
We should also look into:
  • Are there any building code (e.g. in the US) issues related with POF?
  • We should also look at home network architecture with POF? Do we expect a Wi-Fi AP in every room with this? Do we need POF to Ethernet convertor in every room? Do we expect consumer devices (e.g. Laptops and Tvs) to start supporting POF as well?
  • What speed can POF support today and in the next 2-5 years?
Thanks,
Vikas
 
 
 
From: Carlos Pardo <carlospardo@xxxxxxxxx>
Reply-To: Carlos Pardo <carlospardo@xxxxxxxxx>
Date: Monday, September 22, 2014 at 8:20 AM
To: "STDS-802-3-GEPOF@xxxxxxxxxxxxxxxxx" <STDS-802-3-GEPOF@xxxxxxxxxxxxxxxxx>
Subject: Re: [802.3_GEPOF] FTTH Council market study
 

Dear all,

in my opinion, the main advantages of POF in the HN vs CAT 5e are:
    + For "in-wall" installation: POF can be installed in the mains conduit. vs CAT 5e can-not.
        - By law, copper data can not share the same conduit than the mains. (Segurity and EMI / EMC reasons)
        - Introducing POF in the conduit is straight forward, where CAT-5e is typically very difficult. ( flexibility and bending handling )
    + For "out-wall" installation, POF is thinner and easier to hide. ( Under the carpet, etc)
    + Easiness of installation:
        - Connection a POF bare fiber connector is 10 s vs CAT 5-e 5-10 min (RJ-45 crimping).
        - POF and Installation and connection can be done by a lower qualified person compared with CAT-5e. ( Even grandma can do it ;-) )

Regarding costs, POF bill of materials should be similar than copper to be competitive. I do not thing that has to be lower than copper, thanks to the operative cost reduction of POF vs copper.

Regards

Carlos Pardo
        

El 22/09/14 15:59, Dai, Eugene (CCI-Atlanta) escribió:
Hi Serizawa: Thanks for the comments. If we talk about office and /or business applications we certainly have to deal with the comparison of POF with CAT5 cables.  However, CAT 5 cable is not really used home networking although we all have short CAT 5 cables here and there at home. The majority of home networking use either coax or Wifi today. With G.hn products roll out, twisted pair phone line may be used for home networking.  If successful in home networking, POF could be extended to office/business applications.  All that time the points you brought out have to be addressed. If GEPOF PHY is lower in cost than 1000BASET PHY,  than it could compete with CAT5 for that market.
 
This remind me that if we that if we want to bring out the office application for POF, we had a brief discussions at Ottawa meeting, we may have to deal with POF and CAT5 comparison as you suggested. 
 
Regards,
Eugene 
 
________________________________________
From: naoshi.serizawa@xxxxxxxxxxxxx <naoshi.serizawa@xxxxxxxxxxxxx>
Sent: Monday, September 22, 2014 1:18 AM
To: STDS-802-3-GEPOF@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_GEPOF] FTTH Council market study
 
Hello Carlos & Eugene,
 
Thank you for sending information for the FTTH. Also, I looked a material that Eugene presented at Ottawa.
Those information themselves are very good to explain about use cases of GEPOF.  However it can be substituted for CAT 5/6 cables to those applications and it seems to be that they are not explaining about the necessity of GEPOF. In order to convince opponents, we should show them strong impacts and advantages of GEPOF technology. Otherwise, we can't answer if they ask us about it.
We should clarify the advantages / cons. against to CAT5/6 cables (cost, weight, relatability, supply chain, max length, workability, etc).
 
I am pleased you to take in to account the above situations.
 
Kind regards,
N Serizawa
 
 
-----Original Message-----
From: Hayato Yuki [mailto:hayato-yuuki@xxxxxxxxxxxxxx]
Sent: Monday, September 22, 2014 8:23 AM
To: STDS-802-3-GEPOF@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_GEPOF] FTTH Council market study
 
Hello Carlos-san,  Cc Menbers,
 
I understand that the European home network market has been growing more and more.
 
However, we should explain that the POF-cable network is superior to the category-cable network for home networking.
 
Thanks,
 
Yuki@Sumitomo
 
 
 
 
Dear all,
 
please find in this public link:
http://www.ftthcouncil.eu/documents/Webinars/2014/Webinar_27May2014.pdf
the latest information of the European FTTH-Council on FTTH deployment.
 
The FTTH deployment can be used as an indicator of the TAM  for the
gigabit Home Networking market.
The FTTH deployment speed in Europe is around 5 Million houses per year.
 
In parallel with this values, we may add TAM values from ADSL/VDSL
deployment, and new/refurbish homes.
 
Best Regards
 
Carlos