Details of the FCC's Navigation Devices decision
Cable subscribers will be allowed to buy and own their own cable boxes, and connect them to the cable network. An interface being worked out by CableLabs will allow the operator-supplied security element to be separated from the customer-owned digital set-top box.
One surprise is that after 2005, this security interface is mandatory; cable operators will not be allowed to provide boxes with embedded security. Another surprise is that the decision applies to analog boxes as well as digital. But I think this part of the decision just won't be practical. Customer ownership is feasible only because of the revolutionary changes in cable networks resulting from digital video. Here are more of the details of the decision, laced with some of my personal observations.Background
This FCC action springs from a push by Congress for customer ownership of cable boxes. It started with an obscure section of the 1992 Cable Act, dealing with equipment compatibility, but the real decision was made in the 1996 Telecom Act. It happened because Circuit City wanted to sell cable boxes. Circuit City headquarters are located in the Congressional District of Tom Bliley, who just happens to be the Chairman of the House Commerce Committee, the Congressional committee with jurisdiction over telecommunications laws.
The FCC started a proceeding to implement the 1996 law. It became pretty clear right from the start that this was a very complicated subject, partly because the law requires the FCC to avoid taking any action that might jeopardize cable network security, partly because analog cable network designs vary from one cable system to another, and partly because the industry is beginning to shift from analog to digital technology.
And while the law requires the FCC to promote the competitive availability of cable boxes from retail suppliers, it does not give the FCC any authority to require the redesign of cable networks, or to require "portability" so that the same box might work on many different systems, or to require cable equipment manufacturers to license their proprietary technology to competing manufacturers. But the cable industry itself, through the OpenCable initiative of CableLabs, is moving in the direction of more standardization of digital cable systems, including some level of non-proprietary technology and perhaps some level of portability.
The law applies to all multichannel video programming distributors ("MVPDs") except for Open Video Systems. But the FCC decided to exempt direct broadcast satellite systems from the new rules, on the theory that DirecTV, Echostar and Primestar were already selling their set-top boxes, on a nationwide basis. So the new rules apply to cable, MMDS, LMDS and SMATV.
Subscribers will now have the right to attach their own boxes to the cable network. But cable operators have the right to protect against piracy and harm to the network. The FCC took the opportunity to reaffirm that the unauthorized reception of service is prohibited, and the manufacture, distribution and sale of pirate boxes is also prohibited. The cable industry is permitted to adopt standards for "harm to the network." For example, the SCTE might decide to adopt a standard for emission levels from TV sets and customer-owned set-top boxes into the upstream frequencies of two-way cable systems, because some TV sets are designed without adequate filtering.Interfaces
The FCC's goal is the physical separation of security functions from non-security functions. The designs to accomplish this separation must be finalized by the end of 1998, and plug-in modules containing the security circuitry must be available from cable operators by July 1, 2000. For digital security, the Point of Deployment Module ("POD") interface being developed in the OpenCable process in CableLabs is the preferred interface for this separation. For analog scrambling, the EIA/IS-105 Decoder Interface Standard is the preferred interface. Cable operators must begin supplying security modules with the POD interface and/or the Decoder Interface by July 2000, to any subscribers who have been foolish enough to purchase the first-generation retail set-top boxes.
Because this rule also applies to MMDS and LMDS, these market segments may have to accept these same interfaces, even though they did not participate in the standards development process. However, the POD interface was specifically intended for cable systems that transmit network control information in an out-of-band channel. MMDS does not use an out-of-band channel, but carries network control information within each video channel. I wonder what will happen when a subscriber tries to connect an MMDS set-top box to a cable TV POD module?
As a practical matter, I think that you can forget about the analog Decoder Interface. I don't believe that anyone will ever build any boxes with that interface, nor will cable equipment manufacturers ever build any security modules. The interface is just too complicated and too expensive. The physical interface is both a 26-pin connector that looks something like the connector on computer printers, plus an F-connector for a coaxial cable.Separate security
Starting in 2005, cable operators may not deploy any "new" cable boxes that contain embedded security circuits. So from 2000 until 2005, cable operators may deploy boxes that contain either embedded security or separate security. After 2005, the cable operator may redeploy boxes with embedded security if they had been used before, but no "new" boxes with embedded security may deployed.
So 2005 is probably the end of the line for analog security in the cable industry (unless somebody invents a new, simplified analog security interface). While analog programming will still be offered on the basic tier, all premium programming will migrate to the digital channels by that time. The combination of the prohibition against new cable boxes with embedded security, coupled with the costly nature of the analog Decoder Interface, will drive this migration.
The FCC promised that in 2000, after security modules become available, it will reassess the 2005 deadline. These dates are based on a schedule that the cable industry supplied to the FCC, including completion of the POD interface specification by December 1998. But I'm not sure that these dates are really feasible.
While much of the POD specification is already completed, one important hole is copy protection. The very idea of a security interface means that the descrambling or decryption is done in an external unit, and "clear" video is passing across the interface. But the Hollywood studios have become militant against the idea of unprotected interfaces that might allow the copying of their movies, particularly perfect digital copies of their movies. They have required the consumer electronics industry to add copy protection to the interface between digital TVs and digital VCRs. But the particular method that's been accepted for this purpose probably isn't suitable for the POD interface, I've been told by security experts. Without agreement on a new copy protection scheme, meeting the deadline of having a complete POD interface specification by the end of 1998 may be difficult. Missing that deadline means that having security modules available by 2000 would be impossible.
Another possible fly in the ointment is a court appeal. The vote on this decision was 4-to-1, with FCC Commissioner Powell dissenting. His statement, explaining why he voted no, could form the basis for a successful court appeal. He complained that the FCC had previously, in 1996, said that, while cable operators should provide analog security modules, they should also be allowed to offer boxes with embedded security. This 1998 decision in effect overrules that 1996 decision, but Powell says that the FCC did not sufficiently explain why the 1996 decision was wrong. I can imagine the NCTA or a cable MSO going to court and arguing that this decision—overruling the 1996 decision without justification—was arbitrary and capricious. But at most, that would only create a slight delay in an inevitable process.
While these are the major points, there are a number of additional issues. MVPDs must provide information about the network control signals so that the set-top boxes can be designed to process the data or pass it to the security module. For digital networks, the POD interface specification should serve that purpose. For analog networks, disclosing that information might jeopardize network security.No subsidies
While the intent of the law was to allow retailers like Circuit City to sell cable boxes, the law also allows cable operators to sell boxes to subscribers. But the law prohibits a cable operator from using service revenues to subsidize the lease or sale of set-top boxes. The FCC decided that its cable rate regulation policies already satisfy this requirement.
The law gives the FCC authority to grant waivers of these policies. For example, a cable operator might argue that the no-subsidy requirement is a barrier to the development of a new service, and that it should be allowed to subsidize boxes until a critical mass of customers sign up for the new service. But there are no special guidelines for waiver requests. The new rule simply says that the cable operator must show that a waiver is necessary to assist the development or introduction of a new or improved service. You can bet that Circuit City will be looking very closely at waiver requests.What it all means
No, this decision doesn't mean that cable networks will become totally uniform in design, like the PSTN. It doesn't mean that cable customers will be assured of portability, that the same set-top they buy in one city will work in all cable systems. But the trend in that direction is unmistakable. The FCC is pushing toward uniformity, at the expense of diversity, because it thinks that portability would be good for the public. Circuit City is pushing for uniformity, because it knows that this would make retail sales easier. And the trend toward digital technology, with the ability to download applications and network design changes into the box from the headend, is the key that makes standardization of cable networks palatable. The last major change in cable network technology was the implementation of fiber optic transmission systems. That was evolutionary. The next change, digital transmission into subscriber-owned boxes, will be revolutionary.