The FCC’s DVR Rules: Part 1, Connectivity

The FCC’s Notice of Proposed Rulemaking on video navigation devices is an interesting document that with enormous technical and economic implications. It builds in a report by the FCC’s Downloadable Security Technical Advisory Committee (DSTAC) on approaches to passing video content from MVPDs (cable, satellite, and telco-based video distribution systems) to third party DVRs and similar devices.

No Technical Consensus

“Builds on” is a bit of an overstatement because all the DSTAC really did was list some requirements and fail to reach a consensus. DSTAC’s goal was to select some technical standards for passing content from MVPDs to consumer electronics devices sold by Google, Amazon, TiVo and others that would take the place of MVPD-supplied set top boxes and DVRs. It failed to reach a consensus because the standards in question don’t exist. So DSTAC went to the grocery store to buy some items from the FCC’s shopping list, but returned empty-handed because the shopping list included things like cars and chainsaws that grocery stores don’t sell.

When DSTAC –  committee that included bona fide engineers from the MVPD side as well as assorted policy wonks and lobbyists from the consumer rights side – is unable to identify a viable, consensus approach to solving a technical problem, the problem is not ripe for resolution by a rule making. Rather than imposing rules that no one can actually abide by, the FCC should recognize the seriousness of the issue and try to put DSTAC on a more productive track. But that’s not what the commission has chosen to do.

There are two fundamental issues that neither the FCC nor the consumer rights lobby knows how to solve in order for consumers to have access to more video programming on the new devices that have hit the video market in the last three years: connectivity and licensing.

Diverse Forms of DVR Connectivity

Connectivity is the simpler problem and it’s terribly complex. Each MVPD is selling services delivered over its unique network, so it’s never been practical to build a universal navigation device. Cable networks come into our homes over coaxial cables that deliver TV, telephone service, Internet access and some more specialized services such as home security. Telcos also deliver a diverse set of services, but they use different wires (twisted copper pairs, fiber, and wireless) and don’t even all do things the same way. AT&T and CenturyLink use similar technologies, but Verizon, Google, and Frontier use fiber optics. Some wireless carriers use mobile LTE, some use stationary LTE, and some use Wi-Fi. And then there are the two satellite carriers, DirecTV and  Dish, who aren’t compatible with each other.

So each network requires a different approach even to get the content into the device or application that will potentially take the place of the MVPD’s native set top box or DVR. Internet Protocol allows common data streams to move over diverse networks, but only when hardware-to-hardware connections are already taking place. So a universal video box needs to make physical connections to diverse networks to even get off square one.

This only works today when the MVPD supplies hardware that converts its native signal to Ethernet or Wi-Fi. So the attempt to make third party navigation devices without network-specific hardware depends on the presence of some sort of converter box right out of the gate. The devices that currently use CableCard to replace MVPD-supplied DVRs aren’t even this lazy: TiVo, Silicon Dust, and Hauppauge all supply device with native cable network connections today. So the desire to build a universal video navigation device without specialized hardware already takes us a step backward in time and progress.

The issue here is that the current generation of third party DVRs are the functional equivalents of MVPD-supplied DVRs in that they have solved the problem of physical connectivity in the only way it can be solved without littering the home with extraneous devices. From personal experience, I have a TiVo and a SiliconDust device and no MVPD-supplied hardware device of any kind except for the two CableCards that authenticate these devices and decode their data streams. The TiVo has its own multi-terabyte hard drive, and the SiliconDust uses Windows Media Center for its hard drive. Microsoft no longer supports WMC, so I can’t upgrade to Windows 10 until I’m ready to decommission the SiliconDust device. The FCC wants me to lease a device of some sort from the cable company before I can upgrade to its new generation of DVR.

So the connectivity dilemma makes it hard to replace MVPD hardware with third party hardware. Without hardware to connect directly to existing networks, the networks will need to change to accommodate new video streaming devices. This is already happening in various forms for cable, DSL, and fiber optic networks, but it’s hardly complete.

The Carterfone Illusion

Whenever the FCC is faced with an unbundling dilemma, it invokes the Carterfone ruling imposed on it by the DC Circuit Court of Appeals in the 1960s. Indeed, Chairman Wheeler invokes Carterfone is his NPRM comments:

Decades ago, if you wanted to have a landline in your home, you had to lease your phone from Ma Bell. There was little choice in telephones, and prices were high. The FCC unlocked competition and empowered consumers with a simple but powerful rule: Consumers could connect the telephones and modems of their choice to the telephone network. Competition and game-changing innovation followed, from lower-priced phones to answering machines to technology that is the foundation of the Internet. Should pay-TV continue to be an exception? I believe, and Congress has made clear, the answer is no. You should have choices in how you access the video programming you are paying for, as well.

We enjoyed the ability to use modems, answering machines, and cheap phones, but they didn’t get us where we are today. The credit should go to the new networks that replaced the old telephone network because they enabled higher speeds and more device options. It wasn’t cheap phones that got us to the 21st century, it was new networks. Moreover, telephone calls don’t have the intellectual property issues that are paramount in video streaming, so the analogy really doesn’t work. It’s high time we stopped comparing computers to analog phones as an excuse for avoiding hard questions.

Next up: Intellectual Property

Before we get into the intellectual property issues, which I will cover in the next post, we need a much better answer from the FCC about the connectivity problem that comes about from diverse network technologies. And no, Carterfone does not address that question.

It’s not even close.


  • Dan203

    The CE manufacturers on the DSTAC suggested a solution that would work for all technologies. They called it a “virtual headend”, but really it’s a gateway. It’s a box which contains all the hardware needed to access the various services offered by the MSO and convert them to an open standard like DLNA CVP-2 (aka Vidipath). That would allow the retail devices to be designed to a single standard which would work with all MSOs, just like landline telephones. The gateway devices themselves would woldn’t be that hard to make either. They’d essentially have the same hardware as existing settop boxes. In fact some services already use a gateway device. AT&T’s U-Verse uses a gateway to allow their settop boxes to communicate with their VDSL connection. And DirecTVs RVU technology is actually an offshoot of DLNA CVP-2 and works almost exactly like the “virtual headend” the CE manufacturers suggested. The whole thing could be simplified even more for systems like cable where there is a constant 2 way connection because they could push the “virtual headend” off to the real headend and integrating the communication/discovery stuff directly into the cable modem, just like they’ve done with VOIP for phones.

    The real reason the MSOs are pushing back against this proposal is control. They want the “apps” approach because it gives them complete control over the user experience. It allows them to control how the guide looks (i.e. ads) and it allows them to push users to their own VOD/PPV offerings rather then toward OTT alternatives. If you watch some of the videos from the DSTAC meetings this is their main complaint. They don’t want 3rd party devices having direct access to their content without going through their UI first. This is also the same reason the CableCARD 2.0 talks fell apart, because the cable companies insisted on using OCAP, which is basically a Java app downloaded into the 3rd party box that takes complete control over the UI. CE manufacturers, rightfully, believe that the only way they can distinguish themselves is via the UI. If all they are is dumb terminals running the MSOs UI then what’s the point? The cheapest Chinese device on the market is going to win and consumers don’t really end up with a choice.

  • Dan203

    You mention a SiliconDust device in your article. The HDHomeRun is actually a good example of how a gateway device would work. The tuners and security are built into the gateway device and the 3rd party box would access it via a local network. In the case of the HDHomeRun you can use a Windows Media Center PC to record from those tuners, or you can use a DLNA device like a PS3 to watch those tuners live. Neither of those devices need to know how your cable system works, they talk to the HDHomeRun and it does the conversion.

Comments are closed.