Bandwidth Caps May Be Critical Error For Broadband Companies

Technical Writing Geek writes "An Ars Technica article argues that after many years of stagnation, the US broadband landscape is finally 'primed for change'. Companies like Time Warner that decide to cap bandwidth risk being relegated to a 'broadband ghetto. Alternatives to the standard cable modem vs. DSL conundrum will come from technologies like WiMax and (eventually) the 'white space' broadband that might be offered by whoever wins the 700mhz auction. 'All of that is to say that cable and DSL won't always be the only games in town. If wireless solutions are able to deliver on their promises of high speeds with no usage limits, capped cable broadband service like Time Warner has planned is likely to be unattractive, to say the least. Instead of developing plans designed to discourage consumers from feeding at the bandwidth trough, cable companies would be better served in the long run by making investments in new technologies like DOCSIS 3.0 and the kind of infrastructure improvements necessary to meet bandwidth demands.'"

My first first post evern?!

[Demogoblin]

Let them try it, and let the free market decide.

It's about time!

[tecmec]

It's funny that wireless internet access would prompt such a thing though. You would think it would be easier to deliver lots of bandwidth over wires than it would be over the air.

Re:Don't worry, it'll get "better"

[DonCaballero]

But why would cable companies want you to not be able to have unlimited access to streaming and downloadable video?

Oooohhhhhhh, I see what they did there.

I know cable companies are supposed to be evil

[Shivetya]

but its a mighty big assumption that if they offered tiered pricing and didn't see a significant increase in their churn rate that the other guys won't also jump on the bandwagon. Hell people act as if wireless will be the holy grail of internet connectivity convienently forgetting that it is the holy grail for PHONE companies. Getting people to pay for their minutes was probably the biggest cash cow they came up with in a long time with ring tones coming right behind.

Sorry, if Time Warner puts this out and doesn't lose people you can damn well expect it elsewhere. Besides, we don't know what their real pricing model will be, it might be akin to the various levels of slow dsl I am offered by AT&T which ranges from slower than the 80s to almost tolerable - but for download only.

Already an error but not apparently too costly yet

[Average]

The quite independent cable company in this town has always had caps. 10 GB a month on the mid-range $40+ (even more if you don't want video) plan. No cutoff, but warning and overage of $2 a GB (but you can buy add'l GBs at $1 in advance).

I don't typically go over 10 GB. But, I absolutely *hate* worrying about what I've used. So, I live, just fine, on my 2.5/512 DSL line for $25 or so. I'm not even sure why it bothers me. I have no problem with PAYG cellphones.

Lots of people grumble about the caps. But, the cable company is doing just fine. Most people never hit the cap. Those who do are torn between the much-much-faster cable and the hands-off DSL. If they want cable (I'm in the deep minority who would rather have a rooftop antenna than pay $675 a year for TV that still has ads), they'll probably get a cable modem.

It's not about bandwidth from the headend to the home. They can shape that, price that, and build that out. It's about fiefdoms and petty accountants. People who won't sign off on intra-Tier 2 peering agreements because they can't make a buck on it.

Re:It's about time!

[KublaiKhan]

Use the cell tower model and piggyback on existing structures and existing infrastructure. Rent conduit/pole space from the telco to string a bit of fiber--expensive, sure, but a lot less costly than trying to run thousands of last-mile connections. It's not that it won't be expensive--it'll still cost a lot, but there will be significant savings over the traditional wired model.

Re:What is it people have against bandwidth caps?

[kextyn]

Most people who download lots of stuff on bittorrent are already paying more than grandma on MSN. I pay extra for the highest speed I can get. Someone who isn't a geek can get the much cheaper ~$20 DSL or whatever is available. I don't believe bandwidth should be metered and charged like electricity or water. The difference between utilities and bandwidth is bandwidth is not a natural resource that may be non-renewable. If I use 4GB of bandwidth in one day that bandwidth isn't gone, it was just in use for the time I was downloading. That same bandwidth will be available to everybody else as soon as I'm done with it. And how saturated are the backbones anyway? Chances are I can max out my connection all day long and nobody else will notice.

Already available

[delirium of disorder]

I live in a rural (slowly becoming suburban) new development that doesn't have any PSTN (Phone)lines nor cable installed yet. Fortunately, my local ISP [foxvalley.net] offers a wireless solution. It works well, although unfortunately it's based on proprietary Motorola technology [wikipedia.org]. For around $60 a month I get 4Mbit down and 1Mbit up. I have a static IP, and they let my run my web/ssh server from home. I have no formal bandwidth cap, although I was told that if I exceeded 75Gigs a month they might want to talk about upgrading to a more expensive business class connection. The ISP will sell you up to 6Mbit (symmetric) connection per antenna/subscription.

I hope cable companies will make better ...

[porky_pig_jr]

in the long run. Many years ago I worked for BBN and our group was experimenting with the first ever cable modems, that become a "Roadrunner" (still have a t-shirt). At the same time we were looking at DSL (ADSL). Both technologies had their share of problems. In case of cable modem the major issue was the fact that the bandwidth is shared among what's called "Neighborhood Area Network" (NAN), so so protocol had to be in place to insure fair share of bandwidth (as well as user's satisfaction in general). I think at that time that protocol was not the part of DOCSYS, so different vendors had their own protocols (i'm not sure if anything like that is a part of DOCSYS 3.0, or it still leaves the room for competetive advantage). But that was only one showstopper with cable modems. With ADSL things were much worse. Highly sensitive to noise on a line, distance, anything you can think of. Just read about DSL and you'll see why. The best results we had with 'rate-adaptive' modems, those that vary the rate depending on line conditions, so connection was stable, but the throughput was - BLAH. So I remember saying that seems from technical perspective cable modem technology should be more stable vs DSL. Then someone else, more experienced in business of communications that myself said: you may be right, but the cable companies are going to screw it anyway because they have no experience with broadband communications, whereas phone companies do. The guy was right, at least for a while. I still hope cable companies will take advantage of better technology/media they have in place.

(PS. My brother switched from Verizon DSL to COMCAST data over cable a few months ago. He told me the throughput is much better and the connection is stable. He got the whole enchalada from COMCAST: cable, internet connectivity, and voice. Voice is the worst part of the package. He even changed his greetings on answering machine to: "Hello! If you can't get through within the next 30 minutes, call 1-800-COMCAST and complain!")

Re:FP?

[Anonymous Coward]

Profit making enterprises whether public corporations or private businesses are created for the benfit of their owners not customers. This is hardly news. The managers of public corporations have both a legal and I would argue moral responsibility to maximize the returns on investment for their owners (shareholders if you like). Now we can discuss whether this maximization should be over the next two quarters, years, decades or centuries but that is just a matter of strategy to achieve the goal.

Managers of corporations often talk about customer satisfaction, environmental protection, improving employee moral, etc. Such talk is essentially public relations puffery. Now I'm a Socialist but I recognize the logic and efficiency of Capitalism.

The free market isn't always so "free"

[randomaxe]

"Letting the free market decide" won't leave me with a lot of options in the near term. Where I live, the options for broadband service are basically Time Warner (cable) and AT&T (DSL). I currently use Time Warner, and if they start capping bandwidth, voting with my dollar means switching to AT&T... which is like voting "yes" to content filtering.

So, currently, I have a choice between supporting one of two evils, or having no broadband service whatsoever. Awesome.

How Multicast Works with Video Offerings

[billstewart]

Scaling issues are a bit different for a telco DSL service than for a cable network, but they're not _that_ different. A US telco central office typically handles 10K-100K users; a city-wide cable network is similar. Older DSL technology ran the wire back to a DSLAM at the CO, while the newer technology often puts DSLAMs in the green wiring boxes that handle a few blocks so the copper's shorter, and cable equipment typically has a similar-sized box handling 500-1000 users. I don't know which approach FIOS takes. The copper from a new DSLAM to a home can handle about 20-25Mbps, which is about 15Mbps for Television (one approach uses 9 Mbps for one HD channel and 2 Mbps each for three SD channels), and the leftovers are available for Internet data (though some telcos insist on selling it as 1.5/3/6 instead of "whatever's not busy with television right now", which IMHO as a geek is a lame idea about revenue protection.)

In that telco example, the bandwidth limits mean that traffic gets effectively unicast from the DSLAM to the user, because you can't fit 1000 channels of broadcast into 25 Mbps. (By "effectively unicast", I mean it's either actually regular unicast, or it's multicast with only the channels you need on the wire. Same bandwidth etc., just a difference in whether you're in Class D IP address space with multicast handshaking or whether that's all hidden from the home router.) On the other hand, if everybody's watching TV at once, 10K-100K houses at 15 Mbps is 150-1500 Gbps, which isn't realistic. If you feed the CO with multicast, then a GigEthernet can handle about 200 channels of HD or 500 channels of SD, or an OC48 can handle both, and farm it out to everybody who's watching. That's one of the reasons that the telcos want to sell TV as a competing-with-cable service, as opposed to just providing pure transport. (Another is the usual money, competition, etc.)

If everybody's doing typical Internet usage, there are a couple of reasons that the network doesn't melt. The big one is that not everybody's actually burning high bandwidth at once - most of the time you're looking at web pages, maybe pictures, and occasional videos (Youtube etc.), but in practice you can oversubscribe by more than 10:1. Another reason is that TCP reacts to congestion by adjusting transmission speeds and window sizes, so if there are too many people watching Youtube at once, everybody's downloads slow down a bit, but unlike live TV, Youtube doesn't care much.

The other way to get enough bandwidth to the CO is to cache a lot of popular video material there - so either the Akamai model (which is driven by the content providers) or transparent caching run by the ISP (a much older model) can do some of it. It won't catch everything, but I'd hope you could cut Youtube bandwidth demand in half that way.

Disclaimer: This doesn't even *pretend* to be my employer's opinion.

Re:How Multicast Works with Video Offerings

[Shakrai]

I don't disagree with any of what you said, but I don't see a response to my point that multicast isn't a real solution either.

Multicast is a great solution for people that want to watch live TV. It's not going to be very helpful with video-on-demand type stuff though. The majority of the bandwidth that I used for video went into Netflix instant view. Unless there are a bunch of other people on my ISP watching the same video as I am (and pausing/rewinding/fast-forwarding the same as me), how is multicast going to help here? It's a unique data stream for each user, dependent upon what they are watching, when they started watching it, etc, etc. You can bring it closer to the end-user through mirroring/peering arrangements but you can't change it's unicast nature.

Furthermore, while I don't disagree with your assessment of the current "chokepoints" in the typical DSL or DOCSIS network, that's with existing technology. One would like to assume that as demand for bandwidth goes up (be it through VOD, torrents, porn, etc, etc) the market will respond with better infrastructure and more options.

If TFA is accurate then TW's tiers are a pathetic joke. 40GB as the highest one? In a 30 day month that works out to 1.33GB a day or 123.45Kbit/s. Would they seriously have us believe that they aren't capable of delivering more then ISDN speeds (on average) to all of their customers? I think you hit the nail on the head when you said "revenue protection", which is why I'm of the opinion that content providers should not be allowed to be in the content delivery business.

Re:totally naive

[hazydave]

It's true TODAY that most users are consuming moderate amounts of bandwidth (you can't say "average", because most users will consumer average amounts of bandwidth, even if they all go crazy and quadruple their consumption tomorrow). But there are various market forces conspiring to change this. I see this announcement as the cables (and eventually, telcos) hedging their bets.

For example, a top-level Seagate executive announced, at the recent CES show, that "blue laser disc" has failed, and hard drive storage (presumably coupled with downloads) is the future of HD video at home. Let's put this in prespective... if I were to buy a season's worth of any television show on Blu-Ray, that's likely to be around 250GB per season (five discs, fairly comparable to today's DVD set) in the original broadcast ATSC format... divide by two for AVC encoding, forget that division if you're talking top Blu-Ray quality vs. the lesser ATSC broadcast quality, add various bits for audio tracks other than AC-3 5.1, multiple audio tracks, special features, etc.).

I can drive to "Best Buy", buy that, and get home in an hour and 10 minutes, if I know just what I'm after. On download here (HugesNet), that's going to take around 9 days at full bandwidth (1.5Mb/s)... assuming they didn't shut me down due to bandwidth caps, which they would. On a high-speed FIOS link, that would take around a day.

So that day's worth on the 15Mb/s link isn't insane, from the user's prespective. But once everyone's running their internet link at full possible bandwidth for days at a time (basic FIOS would run you three days for this download), well, no one's getting anywhere close to full bandwidth anymore.. the servers at your head end, the local nodes, etc. don't have anything remotely capable of dealing with even a moderate percentage of users sucking down full bandwidth for days at a time.

Truth is, virtually every ISP has a "double secret probation" point, at which your bandwidth is "on notice". Exceed that point too many times, or go totally bonkers beyond that, and you're going to hear from the company. On lower bandwidth connections, such as EvDO, this is a well known means for getting out of your contract without paying a penalty (the satellite folks don't want this, either, so they have the restricted modes spelled out up front).

You're probably talking 5-10 years before there's even the possibility of a real "all you can eat" broadband connection, when viewed through the lens of HD video sales actually replacing Blu-Ray and DVD. As well, you'll need a seriously fat storage means... consider that, at somewhere around 200-250GB per season, my Babylon 5 collection alone, in HD, would eat up the better part of a Terabyte disc. Today, that's an added $40 in storage costs per season or more. Sure, that'll drop, but that's also a 5-10 year thing.

Worse yet, for bandwidth concerns, is the notion you're going to store it all "in the cloud". That would tend to imply that every IPTV home would not become a (15-30Mb/s) * (number of TVs) virtually constant pull, and as well, users will want that to be realtime. Also not happening, and also fully capable of dropping any existing ISP to their knees, today.

We'll talk about the actual whitespace bandwidth (somewhere south of 2.6Gb/s, using current 256QAM encoding and polarization, yielding 60Mb/s per 6MHz slot, assuming zero TV channels and all available for broadband) in any general area. How many clients are expected to be pinned to each tower?

Re:The highest teir should be MAX_BLAST

[Shakrai]

It's not a "constant" stream. I came up with the 209Kbit figure by averaging out the usage over the whole month. In any case, I do the following downstream-heavy activities:

1. Netflix Instant View (45 minute TV episode == roughly 870Mbytes in my experience)
2. Pandora. Pandora by itself will consume 100-200kbits on average while playing, in my experience. If I'm awake then Pandora is probably streaming, as I find it to be a lot better then my local radio stations and easier then trying to manage my own playlists and having the music stored locally.
3. Microsoft (Windows and Office) updates. I fix PCs as a side business. It's easily a couple hundred megs to get a fresh XP install up to date. Add on another hundred megs or so if you also update Office.
4. Other video usage, e.g: The Daily Show & Colbert Report (I watch them on the webpages because I have no cable service), Google Video, Youtube, etc, etc. Not as big of a consumer as Netflix, but it still adds up. This will only grow as more networks put their TV shows online.

There's other stuff I do that probably consumes a fair bit of bandwidth, but it's not stuff I do on a regular basis. A Linux distribution is easily 4 gigs or more, but how often do you download one?

This is not so bad people...

[Deviant]

I am an American who has been living in Australia the last few years. All Cable/DSL services here have usage caps and you pay for faster speeds / more downloads. When I first moved here the idea of the usage caps pissed me off but I have warmed to them. I pay ~US$60/month for 8 megabit down and 384 kbit up DSL with a static IP. I get 24 gigabytes per month and any downloads I do from 1am to 8am are counted at half so it is really more like 30 to me most months. They have a usage meter website that is updated about once per hour and is quite accurate telling me how I am using. If I go over I pay $4/gigabyte.

This is great for a number of reasons. Firstly, everybody has a motivation to do their non-essential torrents etc overnight which improves gaming/voip performance during the day and peak evening hours. Secondly, I have an agreement with my provider where I get such and such amount of data at such and such speed and we are both on the same page - I will never get an email saying to use less and hassling me like I received from Adelphia (now Time Warner I believe) before I left. It doesn't serve as a huge deterrent but it is enough to ensure that you don't waste a precious resource (bandwidth) as readily. If you bought electricity, water, or natural gas on an "unlimited" basis don't you think that would lead to waste as well?

I think that the current "unlimited" system does a disservice to many on a shared-bandwidth medium like cable as well. A few teenagers on a street who saturate their connections 24/7 downloading things like the entirely of the Simpsons etc they may never actually watch make the rest of the neighboorhood slow for things like telecommuting and voip that are much more essential and time-critical. There is no reason/incentive for them to stop or to try to do their larger torrents overnight etc. It is also the shadyness of what the limit really is on the "unlimited" service questions. All in all we can argue about where the pricing and the cap are set but I think the idea is sound and reasonable. They will always let you do what you want but you may have to pay more for a service where you can download 100GB/month than granny pays to do 1-2% of that - as you should.

Re:Bandwidth isn't free, you idiots

[rabiddeity]

Storage is getting so cheap now that it's going to be trivial for the ISP to cache upstream, assuming the protocols are smart enough. Hell, give every customer a box with a terabyte drive and have it autocache whatever is piped to your neighbors. When you select a program, if it's cached close enough to you (or in the process of being cached), then it's no problem. At that point it effectively uses no extra bandwidth, and you could play it at the same time as your neighbor or staggered 6 hours or however you wanted. It would be the ultimate time-shifting device! The point is, how many different programs or movies are people really going to be watching? I'd bet the variety isn't as much as you'd think.

If everybody picked from a menu a few days before and the movie was dribbled over the network in time to be cached and ready to go, that'd be one thing. But that's not the vision. The vision is, "You've never seen Gone with the Wind? Let's watch it now."

I've got 100MBps fiber straight into my apartment in Japan, in what I'd consider to be a rural area. My bandwidth is good enough to grab a torrent of a 2 hour xvid compressed movie in one hour. And then while I'm watching that, I can queue up two more. If the video were streamed beginning to end instead of piecemeal, I could hit play and watch the movie without stopping, but even having to wait an hour it's trivial to invite some friends over, pick a couple movies to queue up, go out to eat somewhere, and have the movies ready to go when we return. The tech is already there, the bandwidth is fine; but the US needs to catch up to the rest of the civilized world.

Re:FP?

[TheRaven64]

This has led to an outbreak of executive leadership that focuses no more than 2 quarters out.

I've proposed a solution to this several times, but to my knowledge no company has ever implemented it. The solution is quite simple. Every year, you pay your executives a reasonable sum of money and issue them with a larger number of shares. The catch is that they are not allowed to sell the shares for five years after they were issued. If the value of the company keeps increasing five years after the CEO leaves, then they will make a fortune. If the company bombs a couple of years after they leave then they will make comparatively little from their last three years of employment.