  Eat Me
join:2002-09-25 Sussex, NJ | reply to cdru Re: 38 Mbps x 8
Verizon's TV service is the same as an 860MHz cable system. If cable companies go to 1GHz and make the nodes smaller, they can provide an equivalent service offering. |
|
  cdru Go Colts Premium,MVM join:2003-05-14 Fort Wayne, IN
| said by Eat Me :Verizon's TV service is the same as an 860MHz cable system. If cable companies go to 1GHz and make the nodes smaller, they can provide an equivalent service offering. Cable cos have to split the 860MHz of bandwidth between television, internet, and telephone service. Verizon has a dedicated lambda for television service and another for telephone/data/VOD. It's NOT the same. |
|
  Eat Me
join:2002-09-25 Sussex, NJ
·PenTeleData
·Future Nine Corpor..
·VOIPo
·Vonage
1 edit | If they go 1GHz they will have 140MHz beyond what Verizon currently has for the TV service. That 140MHz is is 23 channels. at 38Mbps per channel, that's 874MBps. That's a lot of bandwidth.
Throw in MPEG4 or IPTV transport for the TV portion and Cable will have plenty of room to be competitive with what FiOS is offering.
With 1GHz They certainly can offer what FiOS is offering today, and with higher frequencies and smaller nodes, what Verizon is offering 10 years from now.
Fiber may be sexy and all, but coax isn't dead yet by a long shot. |
|
  NetAdmin CCNA
join:2008-05-22
| said by Eat Me :If they go 1GHz they will have 140MHz beyond what Verizon currently has for the TV service. So? FiOS uses a completely separate lamba, or wavelength, of TV service. It doesn't matter if Fios TV is 850Mhz or 1Ghz, it has no affect on the bandwidth available for data, like in an HFC setup.
With 1GHz They certainly can offer what FiOS is offering today, and with higher frequencies and smaller nodes, what Verizon is offering 10 years from now. Currently, yes. In 10 years, no. When you add in telephone, VoD, HD and data altogether in that 1GHz of spectrum, you have at best, maybe five years of parity, but definitely not ten unless you start shrinking nodes down to less than 32 subscribers, at which point FTTH is more cost effective.
Fiber may be sexy and all, but coax isn't dead yet by a long shot. I agree that HFC has a lot of mileage left, but its ability to keep up with FTTH is going to become an increasingly uphill battle over time. -- "This is a bus. You know how big a bus is?" |
|
  n2jtx
join:2001-01-13 Glen Head, NY
·Optimum Online
| said by NetAdmin :FiOS uses a completely separate lamba, or wavelength, of TV service. Or for those on this forum who are technically challenged to various terms, they can add more colors of light to their fiber, each color running at capacity. So I could run TV on blue, Internet on red and Voice on green. If I want more TV capacity I could add orange, yellow and violet. Of course the granularity is much greater than this but the concept is the same. Fiber has much more growth potential. -- I support the right to keep and arm bears. |
|
 etaadmin
join:2002-01-17 Dallas, TX
| reply to cdru said by cdru :said by Eat Me :Verizon's TV service is the same as an 860MHz cable system. If cable companies go to 1GHz and make the nodes smaller, they can provide an equivalent service offering. Cable cos have to split the 860MHz of bandwidth between television, internet, and telephone service. Verizon has a dedicated lambda for television service and another for telephone/data/VOD. It's NOT the same. It actually is »en.wikipedia.org/wiki/Frequency People tend to ignore the real definition of the acronyms. Lambda or wavelength is defined by the speed of light in vacuum divided by the frequency λ=c/f
DOCSIS1~3 have different lambdas for VOD/VOIP TV and internet. 5-42MHz or if you wish lambda 60-7 meters for upstream 88-1GHz for downstream (3 meters-30 centimeters) for internet any 6 MHz segment (DOCSIS1-2) or multiple 6 MHz segments (DOCSIS3) and for VOIP any KHz segments.
In fiber optics you have 1490 or 1310 or else nM (nano meters) λ's
Lambda and lambada are not the same thing  |
|
  Eat Me
join:2002-09-25 Sussex, NJ
·PenTeleData
·Future Nine Corpor..
·VOIPo
·Vonage
| reply to n2jtx said by n2jtx :Or for those on this forum who are technically challenged to various terms, they can add more colors of light to their fiber, each color running at capacity. So I could run TV on blue, Internet on red and Voice on green. If I want more TV capacity I could add orange, yellow and violet. Of course the granularity is much greater than this but the concept is the same. Fiber has much more growth potential. You should know about techncially challened. Doesn't sound like they covered a lot of theory in the technician test... *snicker*
Colors are simply different frequencies. RF or light, it's still the same concept.
There is equipment available today to push cable to 3GHz, which is triple the bandwidth it has today. This is more than enough to offer what FiOS can offer. |
|
 waiting4fios
join:2005-04-08 Howell, NJ
·Verizon FIOS
| Your argument has a few flaws.
1)If cable can switch to a 1 GHZ plant, then couldn't Verizon do the same thing? Also that 1 GHZ plant would ride entirely on the 1550nm video light stream right? That means the 1310 and 1490 still dedicated to voice and data correct? So Verizon still wins. (Same applies to 3GHZ)
2)DWDM = Dense wavelength division multiplexing. Verizon can fit a lot more wavelengths of light onto the fiber but currently it is cost prohibitive to do so in a residential/consumer application, but if consumer demand requires such a route, the fiber has a lot more capacity left in it.
3)If cable were to split their nodes down to 32 subscribers like Verizon has with their FiOS, then why not just go fiber to the home, it couldn't really be that much more expensive right?
4)Doing the math on a 3GHz cable plant with 38 Mbps per channel yields a theoretical max speed of 19Gbps using the entire bandwidth for ip/data (correct me if my math is wrong) Meanwhile Verizon has posted successful trials of 100 Gbps on fiber already.
5)Cable attenuates faster than fiber especially when you get into the higher frequency ranges.
6)In the real world (not lab tests) fiber will work better than coax.
Btw, I do agree however that cable has plenty of legroom left for the needs/demands of today's consumer. But it will never have the same capacity as fiber. |
|
  Eat Me
join:2002-09-25 Sussex, NJ | reply to etaadmin Not only that but the TV portion of the service is almost identical to an 860MHz cable plant. |
|
  cdru Go Colts Premium,MVM join:2003-05-14 Fort Wayne, IN
| reply to etaadmin said by etaadmin :It actually is » en.wikipedia.org/wiki/Frequency People tend to ignore the real definition of the acronyms. Lambda or wavelength is defined by the speed of light in vacuum divided by the frequency λ=c/f DOCSIS1~3 have different lambdas for VOD/VOIP TV and internet. 5-42MHz or if you wish lambda 60-7 meters for upstream 88-1GHz for downstream (3 meters-30 centimeters) for internet any 6 MHz segment (DOCSIS1-2) or multiple 6 MHz segments (DOCSIS3) and for VOIP any KHz segments. In fiber optics you have 1490 or 1310 or else nM (nano meters) λ's What I meant when I said that they weren't the same thing is as you pointed out, BPON uses 3 different wavelengths while cable essentially uses a single block of frequencies. With BPON, the 1310nm wavelength is used for upstream traffic and the 1490nm wavelength is used for downstream traffic. The television portion of the service (excluding VOD) is transmitted over the 1550nm wavelength in the form of a RF overlay. In layman terms, it's broadcasing the same 54-860MHz block of frequencies that cable companies use, using only a single wavelength of fiber.
With all digital channels and compression, BPON-based cable plants don't have to jump through the same technologies that cable companies are having to turn to. Switched digital video (SDV), IPTV, etc isn't necessary (and likely won't be for a long time). That RF overlay is for television, only television, and nothing but television.
For cable companies that are limited within the node to the old 54-860MHz cable plant or the extended 1GHz range, that block of frequencies must be split up between all the services. It's still a finite amount of bandwidth and you can only divide it up so many different ways...and no matter what way you divide it up, it's not getting any bigger.
Fiber has it's costs to deploy a whole new infrastructure, but cable TV companies have signifant costs involved with switching to SDV, IPTV, and/or DOCSIS3 as well. FTTP installations seem to be more ready for the future but cable definitely isn't at it's end of life either. Each side will tell you there technology is better then the other for reason XYZ. There really isn't a "right" or "wrong" technology except one that doesn't work. |
|
  Matt Take me down to the paradise city Premium join:2003-07-20 Jamestown, NC
·North State Commun..
| reply to waiting4fios said by waiting4fios :Meanwhile Verizon has posted successful trials of 100 Gbps on fiber already. That was a long distance backhaul, not to the customer premise. Comcast and Time Warner have successfully trialed the same thing over fiber. |
|
  Eat Me
join:2002-09-25 Sussex, NJ
·PenTeleData
·Future Nine Corpor..
·VOIPo
·Vonage
| reply to waiting4fios said by waiting4fios :Your argument has a few flaws. 1)If cable can switch to a 1 GHZ plant, then couldn't Verizon do the same thing?
Theoretically they can. What's your point? quote: Also that 1 GHZ plant would ride entirely on the 1550nm video light stream right? That means the 1310 and 1490 still dedicated to voice and data correct? So Verizon still wins. (Same applies to 3GHZ)
Voice and data are a still very small portion of the cable spectrum used for the last mile. quote: 2)DWDM = Dense wavelength division multiplexing. Verizon can fit a lot more wavelengths of light onto the fiber but currently it is cost prohibitive to do so in a residential/consumer application, but if consumer demand requires such a route, the fiber has a lot more capacity left in it.
As we've seen, consumer demand drives nothing, at least in America. You aren't going to see what FTTH is fully capable of, at least not within the next 10 years. Verizon will have to get a return on its initial FiOS deployment first. quote: 3)If cable were to split their nodes down to 32 subscribers like Verizon has with their FiOS, then why not just go fiber to the home, it couldn't really be that much more expensive right?
The infrastructure for last mile coax with 32 subscribers each is pretty much there. They probably won't even need a truck roll for most subscribers. It would be transparent and not an 8 hour install for most subscribers. Less cost per subscriber to deploy. quote: 4)Doing the math on a 3GHz cable plant with 38 Mbps per channel yields a theoretical max speed of 19Gbps using the entire bandwidth for ip/data (correct me if my math is wrong) Meanwhile Verizon has posted successful trials of 100 Gbps on fiber already.
Safe to say, you're not going to see 100GBps for a residential internet connection anytime soon, at least not in America. Today's computers would be lucky if they could even use 100 MBps of bandwidth. Most computers still come with 10/100 ethernet adapters, and most gigabit adapters don't even achieve half of gigabit speed under Windows (or Linux for that matter) with desktop hardware. quote: 5)Cable attenuates faster than fiber especially when you get into the higher frequency ranges.
Smaller nodes will overcome this. quote: 6)In the real world (not lab tests) fiber will work better than coax.
I never denied that. quote: Btw, I do agree however that cable has plenty of legroom left for the needs/demands of today's consumer. But it will never have the same capacity as fiber. True. But as we can plainly see, the FiOS fanboys on DSLR just LOVE to play "my provider can beat up your provider." |
|
  Eat Me
join:2002-09-25 Sussex, NJ
·PenTeleData
·Future Nine Corpor..
·VOIPo
·Vonage
1 edit | reply to cdru Re-doing an entire cable network in fiber only will cost more than changing headend gear for DOCSIS3. FiOS' infrastructure deployment costs are already more than the shareholders and investors want to see.
What I am saying in a nutshell that while FiOS can deliver worlds upon worlds of things, you are not bloody likely to see it unless there is a return on investment.
And until you are prepared to pay $1000/month for an internet connection, you will not see much beyond the speeds you are seeing today.
quote: What I meant when I said that they weren't the same thing is as you pointed out, BPON uses 3 different wavelengths while cable essentially uses a single block of frequencies. With BPON, the 1310nm wavelength is used for upstream traffic and the 1490nm wavelength is used for downstream traffic. The television portion of the service (excluding VOD) is transmitted over the 1550nm wavelength in the form of a RF overlay. In layman terms, it's broadcasing the same 54-860MHz block of frequencies that cable companies use, using only a single wavelength of fiber.
Likewise, with cable upgrades, 54-860MHz will be used for one way TV services, 860-1GHz (and beyond) for data and advanced video services. 5-42MHz is used for the upstream.
There are your three wavelengths. It's the same concept as fiber, except that the medium (and modulation) is different.
Throw in MPEG4 compression for the video portion and you can deliver the same amount of video programming in less bandwidth, freeing up more of it for internet access or other advanced services. |
|
 Lazlow
join:2006-08-07 Saint Louis, MO
| reply to Eat Me "Today's computers would be lucky if they could even use 100 MBps of bandwidth. Most computers still come with 10/100 Ethernet adapters, and most gigabit adapters don't even achieve half of gigabit speed under Windows (or Linux for that matter) with desktop hardware."
I do not know what you have been smoking but 100 meg Ethernet is only 12.5megabytes/s most current hard drives are capable of sustained 70megabytes/s with many of the newer(non scsi) drives breaking 100. The vast majority of new computers come with GigE(125Megabytes/s) with a lot of them coming with dual GigE connections. I run both Linux and windows(only when I have to) and have both doing 70megabyte/s transfers without doing anything special. You should also be aware that the vast majority of new machines come with Raid chipsets and that a significant portion of the population is using them( two 70MB/s drives in a Raid0 array will match GigE). Todays desktop hardware is very much capable of handling even GigE speeds. |
|
  Eat Me
join:2002-09-25 Sussex, NJ | You are still NOT going to see gigabit to the home speed in America anytime soon. |
|
 Lazlow
join:2006-08-07 Saint Louis, MO | That may very well be, but do not try to use the hardware(user's) as an excuse for not providing it. |
|
  NetAdmin CCNA
join:2008-05-22
| reply to Eat Me said by Eat Me :There is equipment available today to push cable to 3GHz, which is triple the bandwidth it has today. This is more than enough to offer what FiOS can offer. Here's the problem with that... There is no equipment that supports 3Ghz HFC yet. And for the foreseeable future, there won't be. If you are talking about Vyyo (or another vendor), their equipment is proprietary, which is a huge barrier to wide scale deployment. Not to mention all the plant work that needs to be done to get it 3Ghz capable. -- "This is a bus. You know how big a bus is?" |
|
  NetAdmin CCNA
join:2008-05-22
1 edit | reply to Eat Me said by Eat Me :The infrastructure for last mile coax with 32 subscribers each is pretty much there. Name one provider that is close to having node densities of 32 homes? The average node size is about 500 homes/node.
»www.networks2008.org/data/upload···agen.pdf -- "This is a bus. You know how big a bus is?" |
|
 nanoflower
join:2002-07-14 30876
| reply to NetAdmin I don't see it as that much of challenge once they decide to get rid of their analog channels. Just doing that gives cable plants quite a bit of bandwidth. Then they can start seriously looking at switched digital video. That would free up even more bandwidth. Enough to let them keep for quite some time without shrinking node sizes or moving to 1GHz or higher spectrum usage. |
|
  Jafo232 You Can't Spell Democrat Without Rat. Premium join:2002-10-17 Boonville, NY
·RoadRunner Cable
| reply to Lazlow said by Lazlow :"Today's computers would be lucky if they could even use 100 MBps of bandwidth. Most computers still come with 10/100 Ethernet adapters, and most gigabit adapters don't even achieve half of gigabit speed under Windows (or Linux for that matter) with desktop hardware." I do not know what you have been smoking but 100 meg Ethernet is only 12.5megabytes/s most current hard drives are capable of sustained 70megabytes/s with many of the newer(non scsi) drives breaking 100. The vast majority of new computers come with GigE(125Megabytes/s) with a lot of them coming with dual GigE connections. I run both Linux and windows(only when I have to) and have both doing 70megabyte/s transfers without doing anything special. You should also be aware that the vast majority of new machines come with Raid chipsets and that a significant portion of the population is using them( two 70MB/s drives in a Raid0 array will match GigE). Todays desktop hardware is very much capable of handling even GigE speeds. Your argument is partly true. However, do not forget to consider that most homes now have more than one appliance using the Net.
My home may not be average when it comes to net usage, but I have 8 appliances connected to the home. If I hooked everything up, it would be 10. I can download movies via my DTV on demand, or get them via the xbox 360. Each member has their own laptop, one appliance basically plays music all day, and then there is my office with it's servers and a desktop.
I honestly can see a use for much larger bandwidth TODAY, much less what I will probably need 10 years from now. -- Custom PHP/Perl Development. Vbulletin And Wordpress Mods Too! |
|