The Role of Cable Television in the NII
There are over 11,000 cable television systems in the United States. They, along with the equipment vendors that have traditionally worked with them, represent a springboard to an extremely high capacity, nearly ubiquitous broadband two-way information infrastructure serving both the business and residential communities. This existing resource passes over 97 percent of the homes and businesses in the United States and is connected to nearly 65 percent of the residences. The cable industry has made significant improvements in the quality and reliability of its networks over the past few years and has begun to put in place the structures needed to transform its systems into highly flexible multiple service vehicles. Yet, though we have continually expressed a willingness to make the investment necessary to bring about this transformation of our facilities, there are significant impediments to that goal. The primary barrier to the realization of our full participation in this national information infrastructure is posed by excessive regulation of the cable industry. This federally imposed burden, which undermines our financial capabilities, is coupled with other levels of regulation and barriers to the provision of communication services posed by state regulators and local franchise authorities. When these restrictions are linked with such problems as denial of switch interconnection access and usurious pole attachments and duct rental contract provisions by potentially competitive local exchange carriers, it becomes clear that the most capable and flexible network available today can contribute its full resources to the achievement of a national information infrastructure/global information infrastructure (NII/GII) only if the government promulgates a rational set of regulations that make it possible for full and complete competition to flourish.
The basis for today's cable television networks is coaxial cable, a radio transmission medium capable of transporting a large number of separate radio carriers at different frequencies, each modulated with analog or digital information. It is common practice to use filters to segregate the high- and low-frequency portions of the spectrum to allow simultaneous transmission of information in both directions. This ability to transport coexisting carriers carrying different kinds of information (essentially separate "networks") provides an enormous amount of flexibility and capacity. Today's coaxial systems are, however, arranged architecturally in a broadcast topology, delivering the same spectrum of information-bearing radio carriers (commonly called "channels") to every customer in the community.
The advent of fiber-optic transmission technologies optimized for broadband use during the last decade allows a cost-effective upgrade of existing broadcast coaxial networks to a hybrid fiber coax (HFC) architecture, with fiber trunks providing transmission to and from small neighborhoods of a few hundred homes. This arrangement of fiber and coax segments the traditional coaxial-only transmission plant into many localized areas, each providing a localized assortment of information. When combined with the economics of high-speed digital switching technologies, this architecture allows the simultaneous delivery of multichannel television transmissions and switched video, voice, and data services to and from individual homes within these small serving areas. The upgrade of existing coaxial cable TV networks to a hybrid fiber coax architecture as described above costs less than $150 per home passed and can be accomplished over the period of a few years.
The current design and economic considerations of a hybrid fiber coax cable network call for an initial passing of 500 homes (of which about 300 currently subscribe to cable), but it can be further segmented into smaller and smaller coaxial-serving areas. Its potential digital capacity is at least 1.5 gigabits per second downstream (to the home), and 500 megabits per second upstream. This is in addition to the transmission of 80 channels of broadcast National Television System Committee and high-definition television signals. Networks
configured this way can provide for extremely high-speed access to both symmetrical and asymmetrical data services, as well as the ability to evolve to even higher speeds and higher usage levels through further segmentation of the coaxial "last mile" of the network. Further, because analog and digital carriers using completely different modulation schemes and protocols can coexist, an evolutionary path is kept open for continuing advances in digital services and technologies, while, at the same time, allowing the pragmatic adoption of technology options to build cost-effective systems today.
In building a hybrid fiber coax plant, the cable television industry is providing a cost-effective, reliable, ubiquitous transmission system that can simultaneously support many separate networks, the sum of which makes the necessary upgrade investment a sustainable one for any company in our industry. Time Warner, for example, currently has plans to build at least four separate networks on the foundation of its fiber-upgraded plant. Cable Television Laboratories, an industry R&D consortium, is working with a group of broadband equipment manufacturers to develop an open standard that is called the Spectrum Management Application, to allow the coexistence of these multiple networks and the maximization of the efficiency with which the radio frequency spectrum within the transmission plant is used.
In Time Warner's upgraded plant, for example, there are plans for the coexistence, and separate operation, of networks that will continue the broadcast of scores of analog TV channels and will begin the delivery of high-quality digital telephone service, the provision of high-speed personal computer interconnection service, and access to a wide range of interactive video services. An additional network designed for interconnecting PCS radio microcell sites may be integrated into that residential network, or may be operated independently.
Time Warner's personal computer interconnection services will incorporate support of the Internet Protocol (IP), and they are in the process of working with a number of companies that are designing the necessary hardware and software systems needed to provide a service of this type. Many companies in the cable television industry envision initially offering access to a variety of online service providers, as well as e-mail and direct connection to the Internet at speeds currently unavailable to almost anyone, anywhere. We do not pretend to know how the Internet will evolve; there are those who claim that it will one day provide video, shopping services, and all the rest. It is far from that today and has many shortcomings, as recent security problems have demonstrated. But regardless of whether PC interconnection ultimately flows through a number of competing national online services or through the laissez-faire anarchy of the Internet, cable intends to offer a highly competitive avenue for local residential and business access to any viable service provider.
On a separate front, the industry is just beginning to understand interactive television services through projects like Time Warner's Full Service Network in Orlando, and while this and other trials are beginning to teach us a few hard-won lessons, it is far too early to set the standards for technologies that are still in their early innovative phase. We are, however incorporating standards wherever we can, particularly for content (MPEG, for example.) There will still be public policy questions to be dealt with in interactive TV, but they should wait until the technology and business mature at least to the point that such issues can be intelligently debated based on information and experience that we will gain from the trials that are under way.
An example of the harm that government regulation can do is evidenced by the so-called rate regulations that the Federal Communications Commission (FCC) enforces on the cable television industry. These regulations are over 600 pages in length and are so complicated that the commission has issued corrections and clarifications that total several hundred additional pages. While many people think that these rules are just about the subscription fees that operators may charge to a customer, they do not understand that these regulations also directly affect whether or not we can include the cost of a new piece of equipment or a mile of fiber-optic cable in the cost of "doing business" and adjust our rates to recover the expense. In fact, the top FCC official recently replied, in response to a question about "upgrade incentives," that upgrades are just a code word for rate hikes. Any premature attempts to set either mandated technological standards or regulations shaping business structure have the real potential to slow innovation in a field that may have great future value to American business and society. Such standards may be called for by industries that have a stake in the status quo and wish to limit or confine the directions that innovation may take, but these calls must be resisted if the country is to benefit fully from the fruits of the process of invention and exploration.
In the meantime, the PC networks that the industry is building will follow a much better understood path, since they will use an existing installed base of home computers and their operating software, and an existing communications model as a starting point. Although companies such as Time Warner will be one of many content providers, these will be networks that success in a competitive world will demand be kept open, and they will be as symmetrical as actual usage demands.
The cable industry views the competitive provision of wireless and wired telephone service as an immediate opportunity. The benefits of competition to the American public can only be realized, however, if four legislative and regulatory principles are observed:
The industry strongly supports the concept of universal service and agrees that each telecommunications provider should carry its fair share of the burden to ensure that universal service remains a reality. However, universal service should not be maintained through subsidies flowing directly to particular telecommunications providers. Rather, such subsidies should flow directly to the consumers in need of such support, if the concept of truly competitive service is to be maintained in a multiple-provider environment.
In summary, the existence of a ubiquitous broadband cable television system in this country affords an almost unique opportunity to see the rapid realization of a series of extremely powerful digital networks. Some can offer competition to the existing telecommunications monopolies; some can interconnect computers and computer-based interactive television terminals in ways that can lead to an explosion of new and highly innovative services. In order to open these benefits to the American public, however, government will have to take an active role in lowering barriers posed by incumbent telephone companies and by state and local governments, as well as by current federal cable television regulation. Government will also have to exercise restraint in order to allow innovative and entrepreneurial forces to chart the way into a digital future without the imposition of mandated standards designed to protect a variety of existing interests.
The cable television industry has much to offer in helping this nation to realize the potential of the NII. The other network providers are seeking alliances with us and we with them. If the network of the future is to have a chance to be what its promoters believe it can be, government needs to provide guidance, not hindrance. If it can do that, we can do the rest.
1. For your company or industry, over the next 10 years, please project your best estimate of scheduled construction of new broadband facilities to the total residential and small business customer base, in two-year increments.
Over the next 10 years, we project that the cable industry will upgrade its plant to hybrid fiber/coax architecture with relatively small node areas (approximately 500 homes passed on the average) to virtually the entire cabled universe. Since small businesses are for the most part mixed with residences, these same projections would apply to both bases. Currently, there are approximately 97 million residences in the country, 95 percent of which are passed by cable plant. (Sources: A.C. Nielsen and Paul Kagan.)
2. Over the same period, please project dates when each family of service will first become available, and its subsequent penetration of the total base, again in two-year increments.
The four major families of service which we foresee are multichannel broadcast signals; digital telephony; personal computer network services; and switched interactive multimedia services. Multichannel video services are already available, of course, to all homes currently passed by cable TV plant96 percent of U.S. households. Cable telephony services are beginning to be offered in certain markets in 1995. PC modem services will become available in 1996 and interactive multimedia services are expected to become commercially viable in 1997.
Projected penetration of all U.S. households and estimated market share are shown in Table 1 (availability of these services would track the percentages in #1, from 1999 on):
3. Please outline the architecture(s) which will be used to build this broadband plant.
The architecture which will be used to build this plant will be hybrid fiber/coax architecture, as outlined earlier in our paper, with fiber nodes located in neighborhood with (on the average) 500 homes passed. Based on today's cable penetration, this would mean 300 cable subscribers per node, on average.
4. Please outline the peak switched digital bandwidth (in kbps or Mbps) available to an individual residential or small business user when you launch broadband service, and how that bandwidth can evolve to respond to increased peak traffic and to new, high capacity services (which may not now exist).
Peak digital bandwidth to and from the customer can be examined in the aggregate or on a service-by-service basis. In the aggregate, with hybrid fiber/coax architecture to 500 passings, we must make some penetration assumptions to have a meaningful answer. If we assume that cable maintains its current 60 percent penetration, this results in 300 customers per node. If we assume that 40 percent of those customers avail themselves of digital services of one kind of another, that results in 120 homes. If we further assume that peak usage of digital services is 33 percent (a conservatively high assumption), then the maximum number of simultaneous users being served by a given fiber node at any one time would be 40.
In cable systems as currently being upgraded, the spectrum from 50 MHz to 750 MHz is reserved for outgoing transmissions, but 50 MHz to 550 MHz is assumed to be reserved for broadcast analog services, with 200 MHz remaining for digital services. Assuming 356 QAM digital modulation, with an efficiency of 7 bits/Hz, this results in a total outgoing or "downstream" capacity of 1.4 gigabits/sec. This, when divided by 40 users at peak usage time, yields approximately 35 megabits/sec of outgoing bandwidth available per customer. Even given the expected volume of interactive video delivery, this number is ample for outgoing transmissions. It easily accommodates telephony, PC modem, and interactive multimedia applications.
Incoming bandwidth currently spans the spectrum from 5 to 40 MHz, a total of 35 MHz. Because of noise addition problems in this portion of the spectrum, modulation is probably limited to QPSK with an efficiency of about 2 bits/Hz, yielding 70 megabits/second. If this is divided by the 40 peak users to be fed from a fiber node, this yields 1.75 megabits/second available to each user. This is more than sufficient for telecommunications, including video telephony, and is sufficient as well for easily foreseen applications in PC network and interactive multimedia services.
Both the downstream bandwidth and upstream bandwidth can be increased in several ways. First, the number of customers per fiber node can be reduced on a neighborhood-by-neighborhood basis, based on usage, through the use of spare fibers (which are being installed) to subdivide nodes into smaller neighborhoods. In addition to spare fibers, wavelength division multiplexing can be used on a single fiber to the same end. It is easy to foresee average node sizes of 125 homes passed per node or less, resulting in at least a fourfold increase in the numbers cited above.
There is also the ability to dramatically increase the return spectrum. Time Warner's project in Orlando successfully makes use of 100 MHz of spectrum, from 900 MHz to 1 Ghz, in the return direction. This spectrum has fewer noise problems than the low frequencies cited above, so higher modulation efficiencies are possible. However, assuming the same kind of QPSK modulation used at the low frequencies, this would yield an additional 200
megabits/second of transmission capability, or 5 megabits/second per peak user. Again, this number can be multiplied through segmentation as outlined above.
It is possible to push these numbers significantly further. If very high speed, truly symmetrical capacity is required, frequencies above 1 Ghz can be used. The cut-off frequency of the coaxial cable employed is close to 2 Ghz, allowing for very significant expansion of capacity for high speed symmetrical services.
5. Please project the capital investment you or your industry plan to make on a per home-passed basis to install broadband infrastructure, and on a per subscriber basis to install specific services.
Our experience to date indicates that an investment of between $125 and $135 per home passed is required to upgrade existing coaxial cable television plan to the hybrid fiber/coax architecture referenced earlier.
Assuming a 15 percent penetration rate, we expect the incremental costs per customer moving into telephony to be no more than $1,000 per customer. This investment is largely variable in nature, is made incrementally as telephony customers are added.
It is estimated that PC modem services will cost between $400 and $600 per customer, again, incrementally against only those customers taking the service. This covers the cost of the PC modem, as well as central routers, servers, gateways, and support systems.
It is estimated that interactive multimedia servers will cost between $700 and $800 per incremental subscriber, again accounting for terminal equipment in the home as well as switches, servers, and associated central investments.
6. Please respond to the concerns raised in Vice President Gore's letter (copy of letter attached) regarding the ability of users of your network to original content for delivery to any or all other users, versus the control of all content by the network operator.
The concerns outlined by Vice President Gore are largely addressed in our original paper. We expect to support several different coexisting networks on our broadband transmission system. These range from regulated common carrier-type symmetrical telecommunications services, like telephony, to highly experimental asymmetrical interactive entertainment services. In the middle ground will be a PC network, with great capacity. This network will be as symmetrical as it needs to be, given marketplace demand. As outlined above, we have the ability to expand network capacity in pursuit of the amount of symmetry that makes sense. However, premature installation of capacity and symmetry, in advance of demand, will be prohibitively expensive and, we believe, will not be supported by private investment.
7. Please specifically enumerate the actions which you or your industry believe that the federal government should take to encourage and accelerate the widespread availability of a competitive digital information infrastructure in this country.
We specifically address these points in our paper. To reiterate, they are:
1. All projections (unless noted) are the estimates of the authors and do not represent an official position of the National Cable Television Association or Time Warner Cable.