CableLabs Prepares to Unveil DOCSIS 3.1

forward error correction
that doesn't mean more ping does it?

if not
gimmie dis shit now

Too bad most cable operators will just figure out how to share the extra speed among even more users, without a real increase at the endpoint, yet increase rates anyway.

Yes and no, but it really is the "line noise" that must be worried about. It's always the noise that has to be worried about.

You will be capped at 250GB a month. lol

1000 more people with 15Mbps vs 100 people with 50-100Mbps.

So on the typical ISP with an average of say $30-40 for 15Mbps vs $60-70 for 50Mbps - $99-120 for 100Mbps.

Going with the higher ends here:

15Mbps = $40,000/month
50Mbps = $7,000/month
100Mbps = 12,000/month

I'm no MBA CEO having guy but i think it makes more money for them with lease amount of stress on the network.

I do agree with you however that there should be cheaper HIGHER SPEED plans but a guy can only dream

Low caps, high prices, and high overages... no reason to bother.

no reason? how about SCIENCE!? that's reason enough to keep the research going.

At what point does it become easier just to give every home one big internet pipe and have the tv, internet , and phone go through that instead of splitting everything into channels?

Cablevisions ipad and iphone app already does this . Just bond the channels in such a way that every house gets a symetrical conenction and the cablemodem per se handles all the routing for voice,data,and tv.

as I recall none of the developers talking about bonding 24 - 32 downstream docsis for faster bandwidth using the CURRENT docsis 3.0 modems; so it's highly likely that docsis 3.1 wil use NEW proprietary modems for this new modulation inclusion.

also unknown is can other docsis modulations function on the same coax or must everything else be thrown out to accomodate 3.1? cause if it can't play nice with older modems then EVEYONE's getting the upgrade.. and probably paying for it too.. in the form of higher prices-- not to mention they have to mfg about 30 million modems in the next year-- I suspect they can do this node-by-node rather than the entire network at once so that will help.

How much would faster tiers cost?
It seems reasonable to start somewhere between 150 & 300 megabits for entry level plans. for $30 a month.. if that's the case, 2013's gonna be great provided caps don't ruin the party (or high prices).

Yes OFDM require new equipment at both ends (fiber and coax are still fine) and I believe it can (as currently concived) coexist with D3 on different channels making transistion more gradual. of course they have to hammer out a standard,test it, change the standard, test it. this is a few years away.

Great comment, I agree entirely!

This is a new and unique comment never before posted in this kind of thread.

Thank you for sharing it.

If the ISP's could derail science to avoid ever upgrading their network, they would.

And since we have such robust competition and regulation, who could stop them?

Probably designed to keep one on the $4 month DOCSIS 3.0 modems. Users figures why lay out $100+ at BuyMore and
have hard to sell 'now obsolete' device tho will be 5 years until
TWC even gets to 3.1 land. Not even touch 3.0 really.
Idea I guess is one can always dump the lease modem for replacement upgrade vs eating the full buy. Takes what couple years to break even plus hassle of exchange and no support on
own purchase $100+ Motorola one.
Course to get full speeds there's the big hit to monthly bill when
jumping up tiers.

Companies such as Comcast aren't gonna wait another 2 years, they are already moving forward with FTTP as it's available NOW... the calculus is once google fiber lights up.. Verizon's not gonna wait (very long)to push gigabit plans into the market..

because the transmission medium is Coax, they PHY layer will generally specify a radio frequency mechanism. In the US this is NTSC 6Mhz channels because that is what television used to be. The channels are now commonly modulated digitally using QAM but they are still 6Mhz.These IP based aps are still using cable modems which are DOCSIS based and are likely bonding 6Mhz carriers.

This isn't a bad thing, bonding of smaller rate channels exists within most networking technologies, both optical and electrical, to some degree and as your Cablevision example provides, is transparent to your home network.

Lastly, your Cablevision example is currently being developed in a CableLabs standard known as "Converged Cable Access Platform" or CCAP. It still uses 6Mhz QAM carriers but all of the video traffic becomes IP much like the Voice and Data traffic already is.

See »www.motorola.com/staticf ··· S-EN.pdf for an easy to understand run-down and »www.cablelabs.com/specif ··· 0809.pdf for more exacting technical details.

This is NOT FTTP.
It's to cover the 70% that aren't close enough to the node for the FTTP they offer.
Their current FTTP should cover a lot of small businesses, and in-home offices, and exec,s home account, and those DRIVEN to have the fastest home WOW server , but in the long run if they (remember this is all cable companies AND the same modulation and correction scheme works over fiber and possibly bonded twisted pair ) could offer all sort of users undreamed of speeds, certainly everywhere cable can reach.
and remember how long D3 took even after most decided to skip over D2/D2+ (or whatever it was called) if you see any widespread/major deployment before 5 years I would be suprised, 6-8-10 seems more likely.

Apparently some of us, have selective reading skills or are incapable of change. Fortunately Comcast is cable of changing with the times, and has since increased the caps, and may remove the caps all together for the highest tiers.

They are changing the modulation standard (PHY). Calling it 3.1 seems like it's just a firmware upgrade when this is a significant change.

Aside from the name, if it's cheaper to replace modems and head units vs. pushing fiber all the way to the customer, why wouldn't cable do this?

What isn't clear is how much spectrum is used for the max speed. Is there room for multiple on the same coax run or does the max speed require all of it? If not, how many on a single coax run and how much room is left for QAM video?

If it's possible to design a couple per node + keep the existing QAM video, cable could gradually evolve into a pure IP data and video solution. Most new TV's support IP video streams, as do desktops, laptops, tablets and even smart phones. Could it happen? Boy, I sure hope so. Of course content providers would have to be convinced of this being sufficiently secure.

But I have to ask, is one of the new features of DOCSIS 3.1 lower bandwidth caps and higher prices? I'm really looking forward to a faster connection and returning to Internet browsing with images, gif's, and video's turned off.

Just like the old times that we all love, hate, yet miss.

Cablelabs is developing a channel less standard, well atleast looking into it. IEEE is developing an ethernet over coax standard also.

Anyway with ccap it would allow a much bigger data pipe right ? Since it will free up all the channels for data .

As somebody pointed out above they are working on making it all IP. Look up ccap.

Also if I read all this right t3.1 will allow parts of it. One thing it states is to move up the channels that upstream can use to 250mhz .

This is such a big change, I see them calling it DOCSIS 4.0 when the final adopted standard comes out. The DOCSIS 3.0 name is almost as bad as Windows Vista. Lots of hype and plenty of real world failure, especially in upstream. I see CableLabs wanting to move on from that mess, similar to MS wanting to get away from Vista. Or Chevrolet from the Vega. Or Ford from the Pinto.

As far as using the technology for just adding more users per node, instead of higher speed tiers, I do not see that. I think you will see the elimination of the really low tiers that are only there to compete with ADSL2+. It is silly to see Cox have a 1Mbps down/0.3Mbps up tier in a DOCSIS 3.0 world. It will be even sillier when this new technology is deployed. Offering 24Mbps down/4Mbps up as the bottom tier would be a great way for the cable companies to differentiate themselves from the ADSL2+ telephone companies. They should charge the same constant dollar amount for that service as they do for the 1/0.3 service of today. Just blast the telephone companies with a real value proposition to potential or existing cable HSI subscribers.

Caps will probably increase. I think Comcast and others are going to go the way Cox does. Different caps based on the speed tier you pay for. Plus possibly a standardized overage charge per GB, no matter what tier you subscribe to. The cable companies like the cap clauses. It gives them a relatively easy way to get rid of customers whose activities they may not like. It can be enforced through automatic throttling to dial-up speeds, very large overage charges, or shutting off service.

2 years is possible in competitive markets.. however, all docsis standards have been increasingly long to iron out the UPSTREAM issues.. coax is very good at sending bandwidth signals ONE WAY, but not equally SYMMETRICAL on dozens of frequencies at the same time. as I've said before, it's likely that cable companies will see docsis as too asymmetric going forward.. consumers won't want asymmetrical tiers that create ratios of 10:1 15:1 on bandwidth-- cable companies will have to choose FTTP over coax as the last mile.

cable companies need to be able to get gigabit+ out of the next standard up & down virtually ALL computers made today come with gigabit ethernet ports. it's a shame not to finally take advantage of them with speeds at or above 100 megabits being affordable to the mass consumer.

I agree, this is a first step D4 will probably be the 2 way full speed over Epon EpoC epcot(well it was pretty advanced when it opened ) with legacy QAM and narrowband gone.
OR
whatever beats this for less.
Currently they are aiming at a first, expandable, step that can take over and increase upstream.

as seen no this should be more efficient use of bandwidth then greater then 8 downstream channels per node, and 1/2 the power and cooling at the CMTS.
And it echo on down the line.

Oh how I pine for the days of my dial-up shell account. $30/month, e-mail, Usenet newsgroups and LYNX! Long live LYNX!

Comcast is going to offer 305 Mbps Internet and that will require an 8x4 channel bonded DOCSIS 3.0 modem. With DOCSIS 3.1, they'll be able to offer speeds more than 305.

Cable certainly can handle both directions very well. But the problem is in some areas they need FCC permission to shut down the bandwidth hogging analogs. Yep its their network but the FCC somehow legally gets to tell them what they can stop sending over it.

Optimally to extend the life of Coax will be to go completely switched digital video, But then people come out against the box rentals. Which always makes me laugh because any non OTA alternate to cable requires a full box on every TV.

They will still probably cap the upload at 5m.

Comcast is actually using FTTH for their 1st offering, not coax. It might move to coax down the road though. It's based off their metro-e services right now.