Report By Eric Bangeman | Published: August 14, 2007 - 11:22PM CT

If you're starting to become frustrated with the download speed of your cable ISP or the relatively low number of HD channels available, you ain't seen nothin' yet. Your local cable provider will soon be faced with a serious bandwidth crunch, according to a new report released by ABI Research. The problem comes from more demanding applications, including high-definition video, video on demand, online gaming, and higher-bandwidth Internet applications. The result? Big problems—like fewer HD channels and sputtering download speeds—if the industry doesn't respond to the challenges in front of it.

"The increasing bandwidth demands on cable operators will soon reach crisis stage, yet this is a 'dirty little industry secret' that no one talks about," said Stan Schatt, VP and research director for ABI Research.

Currently, cable providers need close to 750MHz of spectrum to deliver the goods: about 676MHz for downstream applications like analog cable, digital cable, HD programming, video on demand, Internet data, and VoIP service. Upstream spectrum needs are comparatively paltry, totaling about 54MHz.

750MHz of spectrum isn't going to be enough in the future, says ABI Research. "Uploading bandwidth is going to have to increase," Schatt told Ars Technica. "And the cable providers are going to get killed on bandwidth as HD programming becomes more commonplace."

Thus cable providers will need more spectrum to keep customers happy, according to the report, and it may be costly. In order to increase the amount of spectrum available, cable companies will have to make some significant hardware investments, including spectrum overlay devices that could provide up to 3GHz of spectrum. That should give cable companies bandwidth to enable cable providers to offer up to 54 HD channels (instead of just 12), increase its capacity for video on demand, and double download speeds on the ISP side of the business. It's going to be expensive work, especially for an industry that has already spent over $100 billion on infrastructure over the last decade.

Beyond expanding the spectrum available to them, the cable companies have a few other ways to address the upcoming bandwidth crunch, Schatt told Ars. Those include rate shaping (controlling the rate at which data packets are transmitted), digital switching (more on that below), overlaying fiber over existing coaxial cable installations, and better MPEG-4 compression (which will lower the bandwidth overhead for HD video).
IPTV the future

"Digital switching is key," Schatt argues. "Ultimately, the cable companies will have to move to IPTV. They'll be brought kicking and screaming into the 21st century."

IPTV is the technology of choice for AT&T's U-Verse video service. It works by treating video programming as IP data, enabling IPTV providers to send just a handful of channels at a time to a customer's house; traditional cable requires everything the provider carries to be pumped to the set-top box. Schatt believes the bandwidth savings from switching to IPTV will be enough to overcome resistance from the cable companies.

Upgrading to a fiber infrastructure is a much more expensive proposition, and one more likely to occur in areas where the cable companies are facing more competition. It can happen, though—several years ago, Comcast's predecessor on the northwest side of Chicago laid fiber on top of its existing coaxial installation. The payoff is good for both cable companies and users, as it can result in more programming choices and faster Internet access.

If cable companies are up to the challenge, the future looks a bit brighter for cable subscribers. They should be able to get more HD programming, faster download speeds, and better video-on-demand options. But if cable providers are unwilling or unable to address the bandwidth crunch, some of their customers will be on the lookout for Verizon and AT&T installation vans cruising through their neighborhoods.

http://arstechnica.com/news.ars/post/20070814-report-cable-companies-facing-big-bandwidth-crunch.html

i like that article :p

its time to switch to AT&T or Verizon :)

Having channels in digital only and greatly reducing the number of channels in analog will open up quite of bandwidth. Should be able to get 300 MHz by greatly reducing anlog channels. They could get 450-500 Mhz if the get rid of all channels in analog.

Not exactly breaking news.

If at&t can get by with copper pairs, cable will be just fine.

If at&t can get by with copper pairs, cable will be just fine.

They just have to get off their asses and do something about it. Which judging by the TWC situation is harder than it looks for Cablecos.

Bandwidth fairy tales: When will the "Three Little Pigs" get it right?
James Carlini • Published 08/15/07

Just like the "Three Little Pigs" arguing about what they should build their houses out of, some of the incumbent phone companies are in a real quandary as to how they should build out their network infrastructure.

Verizon is definitely sold on fiber. AT&T is sold on some type of copper-based hybrid solution and Quest, along with its new CEO, Edward Mueller, is looking to see what the other two do.

Who’s afraid of the "Big Bad Wolf" of bandwidth demand? This is the one who eats up bandwidth like there is no tomorrow. Every month, he gets bigger and bigger and the Three Little Pigs get more afraid of confronting him. In other countries, he is revered and fed to his heart’s content.

Here in the United States, he is written off as myth and lobbied against as something that will never come to town by "Chicken Little," hired by the Three Little Pigs to scare the legislators into locking in statewide agreements without making the Three Little Pigs commit to real gigabit upgrades.

AT&T is supposedly sticking with some type of copper-based infrastructure. According to the market intelligence firm iSuppli, VDSL (Very High Speed Digital Subscriber Line) is expected to be a mainstream access technology for the next 10 years at least. Perhaps over the next 25 years it will be replaced by FTTH (Fiber to the Home), but for now VDSL should continue to generate healthy sales growth for both OEMs and silicon suppliers.

AT&T thinks the migration path is the right way to go. Digging the land up is the big capital expenditure, but eventually you have to do that anyway to stay competitive.

It is interesting to note that in the last couple of months, according to an iSuppli researcher, AT&T has changed its approach on U-Verse (Project Lightspeed) to now go deeper into the neighborhood with fiber than what it first planned. Evidently, they finally came to the conclusion that 30Mbps wasn’t fast enough. Duh! We told them that in several articles over a year ago. Remember Project Speed Lite?

So now with the use of two twisted pair they can provide 50 Mbps using some new chips. Big deal. In talking with the one researcher, I told him that if I was building warships, why would I go with a better diesel engine to get a couple more knots when I could just put in a nuclear-powered engine? He did not disagree.

Global follower

Personally, I don’t think VDSL on copper is the right conclusion, and based on reviewing global strategies, any copper-based end-user connectivity strategies will put us even further behind other countries that have made the commitment to fiber-based solutions.

When did we fall from being the “global leader to set the direction” to a “weak follower looking at which path of one of the global pack to follow?"

Hmmm, that sounds more like sheep. In either case, a following sheep or a market hog who wants to survive the global slaughter, neither make good decisions when thinking about fighting the inevitable Big Bad Wolf of Bandwidth demand.

Don’t you just love when the “experts” debate questions like, “If we are embarking on an upgrade project now, we should be taking it in incremental steps?” Shouldn’t we be looking at fiber to the premise? Why? Because you can do it once, and be done with it. You can attain gigabit speeds on it that you cannot get with some copper hybrid approach.

It is interesting to read these debates about what copper will evolve to in order to squeeze out another 20-30 Mbps. These experts don’t get the big picture. Forget about competing with COMCAST and other domestic competitors. Set the bar higher. Let them catch up.

We should be implementing technologies that can leapfrog us into the gigabit range to insure a world-class infrastructure. Multi-megabit speeds sound good, but are not what we should be aiming for in future architecture.

Wowing Wall Street

Someone said that they have to be aware of what Wall Street thinks of them. Fiber is a big investment and the numbers might not look good to Wall Street. When did any analyst on Wall Street put in a network infrastructure? And how many were right in predicting the current Wall Street dilemma?

The last time I looked, all the Wall Street experts couldn’t explain all the market jitters that dropped the Market a couple hundred points. They are worried about the credit crunch and rapidly growing foreclosure rate, but cannot connect it with underemployment and job erosion affected by labor dumping. It’s all interrelated.

If they only could see the big picture, they would be saying we need a solid network infrastructure in order to sustain regional sustainability. Economic development equals broadband connectivity, and broadband connectivity equals jobs.

Sorry I don’t buy the “we have to please Wall Street” explanations as to putting in an inferior network design instead of real broadband connectivity. We are losing ground in the world connectivity race.

Will slower speeds and filters deter piracy?

Maybe AT&T wants to deter piracy of video today because piracy really spotlights the weaknesses in their slower speed networks for today as well as tomorrow. According to one article:

Why the big change in AT&T's position against Internet piracy? One reason is that Internet piracy is hogging bandwidth, placing new demands on the Internet's access and backbone networks, and costing AT&T billions of dollars for infrastructure to keep pace with overall demand.

Various analysts estimate that about 5 percent of Internet users account for 50 percent of traffic.

It's also estimated that these users are largely engaging in the free exchange of copyrighted content - movies, music videos, albums, popular TV shows and more - without copyright holders' permission.

More content is moving to the Internet. Knocking out piracy would help ease Internet congestion.

Another reason for AT&T's change of heart is that it wants to partner with the entertainment industry. AT&T is beginning to deploy an Internet-based TV service, called U-Verse, and will need Hollywood's content to build a compelling offering. It wants what Hollywood has, and Hollywood wants AT&T's help in solving the piracy problem.

AT&T's achilles heel

The Internet is showing AT&T’s weak spots as more people move to accessing video content for many applications. Increasing video traffic is showing how AT&T’s approach is not the right one if you want to make it easy to move into more video-based content.

Instead of trying to filter piracy, they should be looking at how they can engineer larger fiber highways, interchanges, and intersections to alleviate congestion on the Internet highways.

Where are all the good engineers anyway? When AT&T had Bell Telephone Labs, money was no object. They could out-build anyone as to quality and speed of a network topology. They were sometimes criticized as being run by engineers, but based on the poor excuse for their new network upgrade, “U-verse,” which has been over-hyped and underdeveloped, those days are long gone.

Is AT&T being run by accountants looking to squeeze every last penny out of that underground copper and lobbyists that want to put in place protectionism of an obsolete business model?

What happened to key management phrases like “build world-class” and “best in class?" Their strategy is perforated with holes of weakness. Who turned executive leadership from a simple strategy of “Build the world’s best” into the “just keep up with the mediocrity of the market” strategy:

“Build what is acceptable, but in no way overbuild or overextend in the various geographic areas that will support a solid revenue stream. This approach shall be approved by the state legislature, but not anything that will cost a penny more than the formula we have applied, which gives strict guidelines to keeping in place obsolete network infrastructure that will not need any significant upgrades or transmission media change-outs until we say as specified in our statewide agreements to provide service that is as good as the competition, but not any better.”

Stopping the Big, Bad Wolf

Sorry but the Big Bad Wolf of Bandwidth cannot be satisfied by copper-based solutions and protectionism strategies. With all the video content being downloaded, the growing amount of games being played and all the other applications like YouTube and MySpace running, he will huff and puff and blow away any copper-based connectivity.

The big fairy tale is not the Three Little Pigs. The big fairy tale is the one being told to the American public that they can compete in a global economy with a copper-based network infrastructure that is designed more to keep incumbents’ pork and profits in place and needed innovation out of their place.

CARLINI-ISM: When you are thrown to the wolves, you either become one, or are eaten.

http://wistechnology.com/article.php?id=4122

Having channels in digital only and greatly reducing the number of channels in analog will open up quite of bandwidth. Should be able to get 300 MHz by greatly reducing anlog channels. They could get 450-500 Mhz if the get rid of all channels in analog.
TWC usually has 75-80 analog channels. Cut out every analog channel, except local channels and about 15 basic cable stations and you're looking at about 250 MHz available just like that. The snag is that they can't just cut channels out of basic analog without renegotiating carriage contracts. But, it'll come in the next few years.

Not exactly breaking news.
Does anyone have the latest on the Lindbergh kidnapping? ;)

what's the difference between IPTV and SDV?

Having channels in digital only and greatly reducing the number of channels in analog will open up quite of bandwidth. Should be able to get 300 MHz by greatly reducing anlog channels. They could get 450-500 Mhz if the get rid of all channels in analog.
Except they have to balance this with ticking off a great deal of analog-only customers who aren't going to take too kindly to suddenly having to rent boxes for all their TVs.

Geez, got anything that is more current than this OLD story??? You can believe what with the new MPEG4 format in combination with VOD technology that all is well...

what's the difference between IPTV and SDV?Internet protocol delivers video as data packets, while switched digital is most often similar to current QAM-delivered video, except it's only sent from local cable nodes when a subscriber selects a specific channel; (then available to all on that node). IPTV can be considered one way to deliver switched video, as table 1 here (http://www.cable360.net/ct/strategy/emergingtech/15264.html) points out. -- John

Bandwidth fairy tales: When will the "Three Little Pigs" get it right?
James Carlini
And that about says it all.

Internet protocol delivers video as data packets, while switched digital is most often similar to current QAM-delivered video, except it's only sent from local cable nodes when a subscriber selects a specific channel; (then available to all on that node). IPTV can be considered one way to deliver switched video, as table 1 here (http://www.cable360.net/ct/strategy/emergingtech/15264.html) points out. -- John

Thanks.

as was said, combine switched digital with reclaiming analogs and there is a crap ton of bandwidth available on your average 750 plant.

as was said, combine switched digital with reclaiming analogs and there is a crap ton of bandwidth available on your average 750 plant.
Not to mention that most systems are designed and built to 860MHz specs so more bandwidth IS available. :)

And that about says it all.
Indeed, good article.

Not to mention that most systems are designed and built to 860MHz specs so more bandwidth IS available. :)
Comcast is already building out to 1GHz in the SF bay area.

Not exactly breaking news.

It's news to me. Thanks to the OP.

Except they have to balance this with ticking off a great deal of analog-only customers who aren't going to take too kindly to suddenly having to rent boxes for all their TVs.
there is that, too. but that's why I suggested a smaller package of analog cable channels. Another obstacle is that it also means a lower price for basic analog, which the cablecos would hate to offer, I'm sure. Soon they can move stations to clear QAM and put in decent guide info so newer TVs w/ built-in QAM tuners will pick up on digital basic channels. Customers with older TVs would get boxes, just like they did with analog boxes back in the day.

"The increasing bandwidth demands on cable operators will soon reach crisis stage, yet this is a 'dirty little industry secret' that no one talks about," said Stan Schatt, VP and research director for ABI Research.

They have been talking about it for sometime. A good start will be to seach DOCSIS 3.0.

Anyway,

http://www.marketwatch.com/news/story/comcast-ceo-shows-off-fast/story.aspx?guid=%7B3074BDBF-2B21-44E7-A6FC-065B42EED653%7D

Comcast Corp. Chief Executive Brian Roberts dazzled a cable industry audience Tuesday, showing off for the first time in public new technology that enabled a data download speed of 150 megabits per second, or roughly 25 times faster than today's standard cable modems.

http://www.multichannel.com/article/CA6414919.html

In an affidavit, Time Warner senior network engineer Ronald E. Boyer said the operator could offer more than 200 HD networks using a combination of technologies, including reclaiming analog bandwidth, improving signal compression, broadcasting MPEG-4 video streams and implementing node-splitting and switched digital video.

It's news to me. Thanks to the OP.

It may be news to you, but it certainly isn't news to the cable companies. In fact the cable companies have talked about this so-called "secret" and what they are going to do about it. It's also been discussed many times in this forum. This new research has not produced new information.

Esser said Cox recently revised its goals for the number of HD channels it would be able to deliver. The operator will get there using a combination of bandwidth-optimization tools and techniques, including system upgrades, retiring analog channels, statistical multiplexing of HD signals and switched digital video, he said.

“I’ve asked Chris [Bowick, chief technology officer of Cox] and his team to find the capacity to offer 50 HD channels by the end of this year, going into 2008,” Esser said. “Whether we do or don’t offer that many channels, I want to have the capacity to do that.”

http://www.avsforum.com/avs-vb/showthread.php?t=864087&highlight=cox

there is that, too. but that's why I suggested a smaller package of analog cable channels. Another obstacle is that it also means a lower price for basic analog, which the cablecos would hate to offer, I'm sure. Soon they can move stations to clear QAM and put in decent guide info so newer TVs w/ built-in QAM tuners will pick up on digital basic channels. Customers with older TVs would get boxes, just like they did with analog boxes back in the day.
But neither of those options would work for many existing analog customers. If they reduced the channels, they'd need to reduce the price. If they demanded the use of boxes, they'd have to be at no extra charge. Finally, most analog customers are unlikely to buy a new TV in order to avoid using a box.

I'm not trying to pick a fight, but it seems odd to me that there's a ton of sympathy going around when someone doesn't want to be "forced" to go to MPEG4 on D* or E*. However, when someone may have to pay an extra $5 or so a month for every TV, or may get fewer channels - that's just business.

My parents are a couple of those analog customers who would have to mess around with making a box work with their analog TiVo and end up with an increase in their bill. For them, it's high enough as it is.

i like that article :p

its time to switch to AT&T or Verizon :)


Yeah - if you can get it.

Not to mention that most systems are designed and built to 860MHz specs so more bandwidth IS available. :)

This is very true, but within that same bandwidth is something called "internet service" that is also competing with the TV bandwidth. You add phone service to the "internet service" and all of a sudden all of that "unused" bandwidth is gone for more TV, SD or HD by the "internet service/phone service" that cable is now making good money off of, so which is more profitable to cable, TV or internet? Right now it is internet and that is where all the extra bandwidth is going, not HD.

so which is more profitable to cable, TV or internet? Right now it is internet and that is where all the extra bandwidth is going, not HD.
110% correct...My earlier post I said that most systems are built to 860MHz specs (1Gig systems ARE out there but they are few and far between) but these 860 systems are only utilizing (still) 750MHz...So there IS still room to grow...not alot but enough for the near future. :)

This is very true, but within that same bandwidth is something called "internet service" that is also competing with the TV bandwidth. You add phone service to the "internet service" and all of a sudden all of that "unused" bandwidth is gone for more TV, SD or HD by the "internet service/phone service" that cable is now making good money off of, so which is more profitable to cable, TV or internet? Right now it is internet and that is where all the extra bandwidth is going, not HD.
I don't think internet and phone service take up much bandwidth at all, a single 6MHz 256QAM slot should be plenty for over 25mbps internet speed, phone, and have room left over for 2-3 SD channels.

Sure, a single person is nothing, but with 300 people on a node all using Internet and phone, it adds up real quick.

Anyone else think cable is fooling themselves with switched digital? I would think a fairly nasty lawsuit would come the day "popular" channels went to switched, mostly from cablecard users. You would think the FCC would be against this, as well.

Anyone else think cable is fooling themselves with switched digital? I would think a fairly nasty lawsuit would come the day "popular" channels went to switched, mostly from cablecard users. You would think the FCC would be against this, as well.
Could you elaborate a bit more please? :confused:

As an individual who has engineered cable systems for over 30 years, let me try to bring some clarity to the realities of bandwidth management.

Traditional cable systems were "broadcast" systems in that the signal originating from the headend was fixed. Everyone connected downstream got what everybody else got.

Fiber overlays allowed cable operators to segment the service areas into much smaller nodes, ranging anywhere from 400 to 2000 homes. Programming can be "customized" to the usage in each individual nodes.

The only thing required to support all the additional demands is to reinforce the fiber backbone and subdivide the service node areas into smaller housecounts. Savvy cable operators have significant "dark" fiber just waiting for activation.

Obviously subdividing nodes takes more than just a magic wand, it takes time and money. But to suggest that a hybrid fiber/coax network cannot support all these new demands is simply ludicrous.

Where I'm going with this:

Any switched service is going to impact:

1) Existing cablecard users, of which many have signed up recently with Tivo S3's, and more to come with PC-based DVR's.

2) Related to the above, but not the same, are users of the new cableco-provided cablecard STB's. These could end up being a white knight, of sorts, as they may force the issue of cablecard usage with switched channels.

As for a lawsuit and the FCC, the users above would have a legitimate gripe regarding their hardware being rendered obsolete by something still supported. Yes, it may seem hollow to those upgrading to MPEG4 DVR's right now, but neither of the DBS companies have pushed MPEG2 as strategic, nor could the FCC care less about how the DBS guys manage bandwidth.

In short, I don't see a happy ending for switched here.

Addendum - the FCC could care less if DBS were singling out "population segments" by dividing bandwidth by demographics. Example here is E* being forced to stop sending "ethnic" channels in a DMA to a wing bird.

And, seeing how most early switched runs have been these same types of channels (although not locals, so a little different), it makes the chance of this ending up badly even worse.