Indirectly, Sanford Bernstein Senior Analyst Craig Moffett argues the crux of cable's challenge is not the capacity of their HFC networks. Rather, it is the legacy contracts with programmers that lock up the majority of their spectrum with analog cable TV channels.
By reclaiming much of that analog spectrum, splitting fiber nodes, and employing switched broadcast video (SBV) techniques, Sanford sees plenty of headroom for cable's service offerings, including HDTV and VOD.
The report argues that this three-pronged approach offers:
...a low-cost path to capacity expansion that keeps cable fully
competitive with even the most fiber-rich next generation networks
(including Verizon’s FiOS network) without a costly rebuild.
Sanford is particularly bullish on SBV. It's common sense, of course. Many of the hundreds of digital broadcast TV channels a cable operator delivers are not being watched at any given time. It is a network inefficiency that can be remedied by simply switching off those channels that are not being watched.
The report notes:
If the idea of an all switched network sounds familiar, it should be. SBV is simply cable’s more spectrum-efficient equivalent of the telco model of IPTV, only delivered over a plant with vastly higher capacity than a telco’s legacy copper infrastructure, and with a logical and low-cost migration path to get there.
SBV is how MSOs can best use the bandwidth they recapture from analog, the report contends:
Each analog channel consumes 6 MHz of capacity (or roughly 1/125th of the total capacity of an upgraded cable plant). Re-claiming that channel slot frees up enough capacity to launch 2˝ more channels in HDTV (each of which consumes about 2.4 MHz), or ten more digital channels (each of which consumes about 0.6 MHz), another ten QAMs, each capable of delivering an additional VOD stream, or another doubling of broadband [Internet] capacity (which today operates out of a single 6 MHz channel slot).
Sanford says Comcast Corp. (Nasdaq: CMCSA, CMCSK) has an aggressive plan to cut the average number of analog channels it carries in half over the next five years, from 70 to 35. That means moving some three dozen basic cable networks from analog to digital-only carriage.
This is where things get interesting:
Until digital penetration reaches 100% – being left off the analog tier means reduced distribution. And that means lower affiliate fees, and lower advertising revenue.
In other words, programmers will fight this tooth and nail. Ironically, it is SBV that may well help prove their case is built on a house of cards. Cable TV programming networks sell themselves to advertisers based on their total distribution footprint – say 40, 60, or 80 million homes. The metric is bogus, as only a small fraction of homes are viewing it. With SBV, MSOs will have all the details, and eventually, so will advertisers.
1. Far fewer customers consume large amounts of capacity than you expect. Dont assume that your usage patterns are reflective of the average user. If you ever saw the numbers you would be shocked!
2. There is upstream capacity. I know an operator that has 60Mb provisioned in the upstream on every node.
3. DOCSIS 3.0 has nothing to do with creating capacity. It just bonds together channels that are available. It provides some efficency from load balancing but its real purpose is to allow a service to be provionsed that is close or exceeds the rate of a single channel (ie. 38M/20M)
You guys who want 100M head over to Verizon, that will pretty much crush their business model because that are making the assumption that usage patterns will be the same for their network. Those assumptions are that users enjoy speed but dont require constant capacity, which is true for almost everyone...
re: I wish! The fist carrier to do this will have the future network, and my business.
During the early days of public internet access, when a competitive common carriage network was made available to entrepreneurs, there were something like 4000 or more independent ISPs. Today how many are left? If there were a real business in fat access pipes to residences at least some of the 4000 would have tried it and proved it out. Hasn't happened.
The reality is, while many people will overpay by thousands and even hundreds of thousands of dollars for an abode, few will contribute a nickel back towards common infrastructure that benefits all. Selfishness wins in the US even when the cost is shitty infrastructure in return.
So if we want a fat pipe we'll have to buy bonds issued by a broadband utility and we'll need to make monthly payments sufficient to pay back those bonds, plus interest. Unfortunately, some of the proceeds raised by the bonds will have to be spent on fruitless activities such as legal expenses fending off lawsuits by the incumbents. Consider that just an additional tax imposed by goverment agents claiming to be acting in our best interests :-(
> why not just sell a fat, open IP pipe to > the home and rattle the entire pay TV > business from the ground up? :)
I wish! The fist career to do this will have the future network, and my business. That's all I want, is a freakin' fat IP data pipe. Voice, data, video and any other services will ride on this. My cable companies requires a "bundle package" - I can't get only data service from them. If I could, I would be a customer. I can add the voice, video, etc on that in 10 minutes.
The only way for VZ or cable companies to survive in the long run is to recognize this, accept it, and start designing add-on IP services for their customers to suplement data-pipe revenue.
re: Arguably, this is a boneheaded move by VZ. Why deploy a me-too service with the associated video content cost model?
The only reason I can think of is to feign facilites based compeititon in order to minmize regulation. Also, I'm no insider to the FCC but isn't Commissoner Martin buddy/buddy with some high level VZ execs? If so and Martin is an idealist trying to change the world by government fiat he may have traded with VZ, promising both state and federal deregulation for VZ promising to deploy access fiber. If so, VZ has to look like it's really trying w/FiOS else Martin realizes he has been conned.
re:For VZ, why not just sell a fat, open IP pipe to the home and rattle the entire pay TV business from the ground up? :)
VZ makes more money by keeping bandwidth scarce and charging accordingly. Also, fat and open pipes are expensive, have huge sunk costs and will drive VZ's primary revenue source, voice, towards zero. Shareholders would revolt, as they should, if VZ really did this.
RJ, thanks for the lively discussion on this topic. Back to your earlier point....
I think you're missing the point. If MSO's open a wide pipe to MSO households in a nondiscriminatory manner they will be bypassed.
MSOs are keenly aware of this risk, that's why it will never happen in the U.S. in the near term.
The same goes for Verizon, now that they've essentially set themselves up as competitive cable TV providers. Arguably, this is a boneheaded move by VZ. Why deploy a me-too service with the associated video content cost model? Cable moved into high-margin data and voice to supplement their core video business. For VZ, why not just sell a fat, open IP pipe to the home and rattle the entire pay TV business from the ground up? :)
re: "In the bypass scenario MSOs get to start playing the Cost of Bandwidth games, charging heavy traffic sources (e.g. HBO, ESPN) for access to their networks (and subscribers)."
Whose subscribers are they really? To HBO and ESPN (or should I say Time Warner and Disney) they are their subscribers and not the MSOs. The MSO is merely a distributor for their content. Take away the content and MSOs go bankrupt very quickly.
Consumer QoS is a red herring. It'll never work.
Consumer hosted servers probably won't work either. It's cheaper and easier to host at an ISP. It scales better too.
Just because cable's primary expense is programming and that the media companies hold the stronger position, both of which I agree, doesn't mean cable companies are selling at a loss.
On the CPE. It's in the cable companies interest to control this device just like AT&T did in the early days of the telephone. The problem with this model is that now that the MSOs have to earn a return on these STBs if they are to meet Wall St. expectations.
An alternative financial model is that the CPE is purchased by the consumer (and effectively immediately depreciated). And hence no ROA required by Wall St. I think that's how radio used to work during the days of RCA. Also, itunes/ipod is another example. Time Warner et. al. would have to sanction the devices and consumers would have to want to buy them. I believe consuemers would as long as they offered new bling as consumer's like to buy electronic gadgets.
Of course, MSOs do have a huge advantage due to the sunk costs associated with an OSP. (Note: I find it interesting that the aggregate phone system's PP&E is valued at more than those of the cable cos. It suggests these financial models, reported to Wall St. by the incumbents are distorted, though I digress.)
Anyway, somebody would have to solve the problem of sunk costs in order to build a real FTTH system. That's a long shot, but it's a real problem and one the US needs to figure out.
Michael Harris asks: Alchemy, are you suggesting MSOs are disadvantaged versus DSL-based architectures, i.e. that which is being deployed by every major U.S. telco except Verizon?
Not at all. As I said in my last post, you can keep doing node splits and offer huge amounts of bandwidth to each customer once DOCSIS 3.0 is deployed. The fewer subs you have hanging off a fiber node, the cleaner the plant so you can also add in more exotic modulation than is specified in DOCSIS today. You can also take advantage of the higher frequencies supported by DOCSIS 3.0 since each segment is so small.
Yes, few MSOs are making money from media over IP, but they are making billions delivering it. That's their business. The same is true in video, even in their traditional cable business, MSOs are primarily distributors of content.
What I'm saying is that MSOs are very conscious of theft of service. Today, they have huge numbers of people stealing basic cable. The last thing they want is to provide a big, fat, unpoliced upstream pipe so their subscribers can pirate the MSO video feed and distribute it over the IP network. You can do that today with slingbox. It wouldn't be hard to take the HD digital set-top box output, stuff it in a cheap PC, and fire it out to a cable modem. The only thing stopping you is the relatively low upstream data rate. You couldn't do it today in real-time but you could certainly capture content in a file and stream it with BitTorrent.
What competitors are better positioned to *deliver* broadcast and on-demand video content? Satellite? AT&T? Not. VZ? Maybe.
I think the cable operators own this space. It will be much more difficult for the ILECs to crack into the video distribution mass-market than it was for the MSOs to get into the residential telephone business. It all boils down to the huge cost advantage of already having a COAX wire running to every house. You can swap out the infrastructure behind that hunk of COAX relatively cheaply compared to the ILEC expense of running fiber to every home.
degsyw claims: As you say MSOs need to limit capacity as the latest applications would ruin their network. Why? because there is so little of it, need I say more.
traffic shaping .... fooey ... traffic shaping policy is throw away valuable customer data when he has already sent it down the wire .... as you say they need a crisis so they can throw more traffic away.
I'm saying that the MSOs will need to traffic shape the upstream to slow down all that peer to peer traffic. If you had the capability, you'd do it in the cable modem. Unfortunately, all you can do on cable modems is port block. To shape BitTorrent or Slingbox, you need deeper packet inspection that will have to be implemented as an ajunct to the CMTS.
They cannot meet the demand unless they change the way they do business, if you told me they were going to use the fibre and duct in their network to push ethernet switches close to the customer premises and open up bandwidth to meet demand you would get my agreement, telling me that a failing broadcast industry is going to restrict capacity of new applications to save their necks, think again.
Why would they ever want to do that? There's already plenty of bandwidth on the cable plant. The issue is how much an operator is willing to spend doing node splits. You're never going to have one fiber node per customer. As data requirements grow for the true heavy hitter, HD Video on Demand, the sweet spot might be more like 50 homes per fiber node. That's much cheaper than an ILEC running fiber to every home. At 50 homes per node and with DOCSIS 3.0 upstream channel bonding, you can make the upstream as big as you want. I claim nobody will do that since they would never want their customers to be able to steal and redistribute content in real time.
"This argument only holds water if MSOs are selling HBO at a loss. If they have any margin then HBO could sell direct and cut them out. Of course, this wouldn't happen overnight and the transition would be messy, but that's the end game."
Didn't we all discuss this in about a dozen different IPTV threads over the past year? In the bypass scenario MSOs get to start playing the Cost of Bandwidth games, charging heavy traffic sources (e.g. HBO, ESPN) for access to their networks (and subscribers). Maybe billing the subscriber for the inbound traffic from such high-volume sources. Threatening QoS shaping on the bypass traffic and claiming that it adds value to the customer (e.g. by "guaranteeing QoS" of the locally sourced traffic).
But upstream capacities also play into revenue erosion: as bandwidth increases, more subscribers can viably self-host applications which otherwise must be hosted by an ISP. Will it lead to the likes of a "MySpace" page or perhaps a NannyCam hosted on your machine at home? Who can say. But let's not lose sight of the question: How do I profit from it?
The blogs and comments are the opinions only of the writers and do not reflect the views of Light Reading. They are no substitute for your own research and should not be relied upon for trading or any other purpose.
To save this item to your list of favorite Light Reading content so you can find it later in your Profile page, click the "Save It" button next to the item.
If you found this interesting or useful, please use the links to the services below to share it with other readers. You will need a free account with each service to share an item via that service.
Digg
Del.icio.us
Reddit
