Technology, investments fuel network deployment

A year-end surge in phone company commitments to installing broadband networks has radically altered the “digital superhighway” landscape. It now looks as if the high-speed digital communications infrastructure will be rolled out much faster than almost anyone thought. And — big surprise — it looks as if it may be the phone companies rather than the cable companies that set the pace.

Pacific Bell opened the floodgates in October when it announced that it will replace its entire existing copper wiring with a new hybrid fiber/coax network (see Vol. 3, No. 7, p. 3). In early December, Bell Atlantic announced an even more aggressive network development schedule. Officials at Nynex and Ameritech tell us that they are on the verge of naming vendors for major deployment schedules as well.

Together with the full-service network schedule announced by US West early last year, these commitments represent the lion’s share of the telephone industry. The only major carriers still to be heard from are BellSouth, Southwestern Bell and GTE.

Thus, while all the attention has been focused on the upcoming cable company trials and on mega-deals between telcos and out-of-territory cable companies, the telephone companies are upgrading their existing plant to support video dial tone and other interactive services. These telco plans for implementing full-service networks dwarf what has been announced by TCI, Time Warner and all the other cable operators combined.

All told, the total regional Bell operating company commitments call for high-speed connections to be available to 800,000 households by the end of this year, 4 million by the end of 1995 and 9 million by the end of 1996. Given the cost and revenue considerations that underlie this explosion in telco activity, it’s reasonable to assume the as-yet-uncommitted local exchange carriers will soon adopt similar rollout schedules. Allowing for a year’s lag time for the late bloomers, this would make high-bandwidth service available to as many as 4.7 million households by the end of 1995 and 11.2 million by the end of 1996.

This does not mean, of course, that all of these households will be using high-bandwidth services — or that all of the servers and services will be in place — merely that the wiring is available to this many homes. However, no matter how you look at it, this would be a staggering achievement.

SOUND ECONOMICS FOR TELCO NETWORKS

It was not that long ago that the phone companies were telling us that it would take decades for this to happen. And it was less than a year ago that the phone companies were touting the benefits of ADSL (asymmetrical digital subscriber line) video services over existing copper wires. What happened?

Incremental deployment costs. One answer is that the telcos are finding that the new hybrid fiber/coax systems will be so much cheaper to install, maintain and operate than the existing copper-wired system that they can just about justify replacing their existing plant — even for existing telephone and data services.

PacBell, for instance, says that if its fiber/coax system were to be used only for narrowband communication services, it would be 36 percent less expensive than the traditional copper plant to install and 20 percent less expensive to operate. Of course, most of the areas targeted for deployment of this broadband network are already served by copper plant with plenty of life left in it. But even here, the cost of the existing wiring is depreciated over the expected life of the wiring. A percentage of the cost of the new wiring can be charged against the fact that the existing wiring would have to be replaced at some point anyway.

On the operations side, PacBell says its cost of adding new services or connecting new customers will be 76 percent lower with the new automated network, and that it will realize an annual savings of 78 percent in maintenance and repair, representing close to $10 per customer per year. In all, the company says the annualized value of these cost savings comes to $50 per customer.

These new economics suddenly make replacing the existing phone system an enticing proposition. It will not take a lot of revenue from new services to make the new system hugely profitable. For example, for the 360,000 households to be served by the new PacBell network in Los Angeles, the capital cost allocated to establishing the ability to deliver video dial tone services comes to slightly more than $46 million, or $128 per household. It is not clear yet that the Federal Communications Commission will sign off on the cost allocation between “old” and “new” services that yields this figure, but this is a tiny fraction of any previous estimate of the incremental cost of providing interactive broadband services to the home.

ADSL in the interim. The result is that the telcos now regard ADSL transmission of video over existing copper wiring as a short-term, stop-gap measure that will allow them to get some test-market experience in the coming year before the fiber/coax systems are deployed.

REGULATORY CLIMATE FAVORS TELCO MODEL

The phone company projections are, of course, just that — projections. No one knows what the demand will be, how fast it will develop, or what the legal/regulatory environment will be. However, in the wake of Vice President Al Gore’s January 11 policy statement (see p. 8), it looks as if the climate can only get more favorable, rather than less favorable. Any loosening of the regulatory environment should help to accelerate the time schedule.

But even if the systems are rolled out at half the pace projected, they are going to create a huge potential market in just the next three years.

JUST HOW OPEN IS ‘OPEN ACCESS’?

Equally important, this market starts from a very different place than does the cable market. The telco market is — and is certain to remain — a “common carrier” market, with “open” access to all comers. Given these new telco plans and the Administration’s commitment to open access, it is a pretty safe bet that telco common carrier rules rather than the old cable company “gatekeeper” rules will be the model for the future. However, there are likely to be some real, practical restrictions on just how “open” the systems really are.

Gateways. Video dial tone and other services for telco systems are to be provided via what the FCC calls “gateways” established and run by service providers. The phone company itself operates the first-level gateway that opens up access to any number of unregulated “level 2″ gateways through which both traditional and interactive services can be accessed.

The phone company may also (and probably will) operate its own level 2 gateway. Other companies may also operate level 2 gateways — as long as these conform to the interconnect standards established by the phone company (which could turn out to be different for different phone companies). Someone who wants to offer services over the network can set up his own gateway or sign up with the phone company or a third-party gateway provider.

The phone companies are setting up their gateway companies as separate entities that (supposedly) deal with the phone company on the same basis as any other gateway operator. In reality, because the phone company and the phone company’s gateway subsidiary will set the standards, it is likely to be some time before there is real, serious competition in the gateway business.

The PacBell filing makes this abundantly clear: In order to achieve compatibility among services and thereby to make all services readily available to customers, PacBell will require that “all technology and selection processes that are used by service providers will be compatible with Pacific Bell’s video dial tone network to facilitate and simplify users’ interactions with the gateways.”

In other words, all services must flow through the compatible hardware and software systems. The household network interfaces and user terminal operating systems, APIs (application program interfaces) and other protocols will be set by the telco. The market will be “open,” but anyone who wants to play must duplicate what PacBell has chosen as its standards.

Channels. You may note that different telco systems claim different numbers of “channels.” As we will discuss below, this has more to do with system architecture (how many homes are serviced by a common fiber-optic feed) than with the service provided. All of the systems will provide a large number of analog cable channels plus a dedicated “channel” for each live interactive user. The interactive content will move over this dedicated channel.

Settop boxes. Phone companies will have to allow for open competition in the user terminal or settop box market. However, as with the gateway servers, the phone company will establish the network conventions and protocols along with program interface and operating system standards for programs that run in the user terminal boxes. Anything attached to the network must be compatible with these specifications. There is, at present, no mechanism to ensure common standards between systems.

Additional advantages. The FCC rules offer the telcos another major advantage in setting market conditions by providing it unregulated freedom to offer a variety of support functions in the video dial tone arena. These include marketing activities, order taking, billing and collection, home wiring installation and maintenance, and the provision of video conversion equipment, such as settop boxes and remote controllers.

In the case of Pacific Bell, the company says it plans to establish a non-regulated affiliate as the service bureau that will handle these functions. While anyone who wants to can set up competing service support systems, the telco is clearly positioned to have a tremendous advantage, which is likely to translate into setting the tone for all aspects of the business related to these functions.

The carrier will also have the ability to include non-regulated, non-video services that it owns and provides as part of the package accessible through its gateway and to retain those services exclusively for its gateway. These might include audio services as well as enhanced information services.

BROADBAND COMMUNICATIONS AT BARGAIN BASEMENT PRICES?

All of this adds up to further inducement, beyond direct transport revenues, for moving forward with deployment of broadband networks. But there’s an even greater incentive that’s built into the way the telcos are regulated, which relates to the vital issue of cost allocation and pricing.

As noted earlier, PacBell figures that most of the cost of building the new system can be allocated to existing phone services without increasing the rates for standard service. If it convinces the California Public Utility Commission and the FCC to go along with this reasoning, it will be in a position to allocate very little of the overall cost of its new networks to video dial tone. This would mean the lion’s share of construction costs would be covered by the regulated revenue base of telephony operations, lowering the financial risk of broadband deployment and leaving the telco room to price video dial tone services aggressively against cable and other competitors.

Pending state, federal OK. It remains to be seen whether such cost allocations will fly at the FCC or at the California PUC. But chances are they will, and, if they do, the telephone industry will be well on its way to winning government approval for getting into broadband communications at bargain basement rates. As a result, transport costs to providers of advanced digital services could prove to be far lower than anyone dreamed they would be, making it much easier for digital media to penetrate the mass marketplace than once seemed possible.

WHO IS DOING WHAT TO WHOM — AND WHERE?

Telephone companies have been applying to the FCC for permits for video dial tone tests for the past year. Now, even before most of the tests are under way, they are applying for approval for commercial operations. At press time, phone companies had applied for authorization to begin construction of video dial tone facilities for service areas that encompass 2.4 million households. A brief rundown on who is doing what follows.

PacBell. Pacific Bell laid out its initial deployment plans in a year-end filing for video dial tone approval involving 1.3 million households in four operating territories, including southern parts of the San Francisco Bay area (490,000 homes), parts of Los Angeles (360,000), several Orange County, CA, communities (210,000) and San Diego (250,000). The plan calls for construction to begin in the second quarter of this year, with all communities to be online by the end of 1996.

As we reported earlier, the new system will use fiber-optic lines running to each neighborhood with a coaxial cable “bus” carrying service to (and into) the homes in the neighborhood. In the PacBell system, each fiber “node” will service 480 homes. Each of these areas is further broken into quadrants, so that only 120 customers are on any given coaxial “bus.” Each quadrant has access to the full range of bandwidth available on the coaxial cable, which includes all the analog channels (70) as well as the bandwidth allocated to digital services. (See Vol. 3, No. 7, p. 3, for a discussion of the PacBell network design.)

Within the digital bandwidth an individual user “channel” is established when the customer tunes the settop terminal to the digital services. This sets up a dedicated link between the central office “miniheadend” and the customer. The actual data bandwidth allocated to this “channel” will vary to accommodate the content being transported. A movie, which can be compressed in advance, might require half the data rate as a live sports event that requires on-the-fly compression and lots of camera movement.

On January 24, PacBell announced that it would deploy video-on-demand services to 100,000 homes by the end of 1994. Hewlett-Packard will supply video servers for the system. No other hardware or content providers were named.

Bell Atlantic. In December, Bell Atlantic filed an application with the FCC for rollout of a video dial tone network across a 250,000-household territory in northern Virginia and Maryland outside Washington, DC. The move, which envisions use of existing copper loop as opposed to installing fiber and coaxial links, comes on top of previous filings calling for fiber star networks in northern New Jersey extending to 50,000 households.

All told, Bell Atlantic chairman and CEO Ray Smith says the company would have 250,000 connections in place by the end of this year, with a million more to follow in 1995 and 1.5 million per year thereafter through the end of the decade. While some of the market base will be served by existing copper in the short run, officials say the company will move to higher-capacity networks like those slated for New Jersey very quickly in the years ahead.

US West. The latest filing came in early January as US West earmarked Denver, Portland, Boise and Minneapolis-St. Paul for broadband facilities connecting some 750,000 households. Construction on the networks is to begin this year, while the already authorized 60,000-home trial in Omaha is still in progress.

Nynex. Nynex and Ameritech are set to be the next RBOCs to announce major broadband rollouts. Sources at both companies indicated the initial hookups would be in the range of 200,000 to 300,000 homes each by year’s end or early in 1995, with deployment accelerating to an annual level of 1 million connections by 1997.

Joe DeMauro, Nynex director of new product development, refused to name sites but said the initial thrust would concentrate on areas where the company felt most threatened by the likelihood of accelerated expansion to full-service capabilities by cable companies. In metropolitan New York alone the telco faces competition from Time Warner/US West, Bell Atlantic/TCI, Bell Atlantic/Cellularvision, Cablevision Systems Corp., Teleport Communications Group and MFS Communications.

Nynex was set to launch a video dial tone trial in New York City sometime in January, but according to DeMauro, broadband deployment will begin before the trial results are in. He says the company will pursue network design strategies that require deep penetration of fiber, although the firm is likely to use existing copper plant for some early iterations of video service, much as Bell Atlantic is doing.

Ameritech. Ameritech’s plans are still under wraps, but officials disclosed that the company is on the verge of unveiling a broadband deployment schedule that includes two immediate 100,000-home projects in Naperville, IL, and Troy, MI. It expects to start construction on both this year. Sources say Ameritech will use the fiber/coaxial “star/bus” design common to the cable industry and recently adopted in distinct permutations by Pacific Bell and US West.

Although Ameritech, like most other RBOCs at this point, has filed a petition in federal court seeking overturn of the federal cable-telco cross-ownership prohibition, spokesman Mike Brand says the company intends to go forward with video network deployment no matter what the court decides. A federal district court in Virginia has already ruled that the cross-ownership ban is unconstitutional as it applies in Bell Atlantic’s operating territories, but it refused to apply its ruling to other telephone territories.

While entertainment video services will be a vital part of the service package to be delivered over the new Ameritech networks, the company intends to exploit opportunities in non-entertainment areas as well, Brand says. He noted that, even without further freedom to participate in video entertainment programming, the telco has court-sanctioned freedom to develop and provide information services, including non-entertainment video.

The company has been experimenting with such services in a variety of tests and believes market demand in this arena will be strong enough to contribute significantly to the new networks’ revenue streams. “Entertainment is part of the pie,” Brand says, “but we see opportunities in health care, education, work at home, work skill enhancement and other areas which we’ve been testing.”

Along with deploying fiber/coaxial hybrid networks, the company, like Bell Atlantic and Nynex, will also make use of ADSL technology to provide video dial tone over twisted-pair copper lines, says Brand. “We see this as an interim technology, but it should be very useful,” he added, voicing sentiments held by Bell Atlantic and Nynex as well.

‘Star/bus’ — Ameritech style. Ameritech is taking a unique approach to the star/bus design. Most cable and telephone planners envision coaxial serving areas of about 500 households per dedicated fiber link. Ameritech, by contrast, appears to be leaning toward 2,000-home serving areas, where four separate coaxial “bus” systems would operate off the single fiber feed to accomplish the same level of user-to-bus-capacity ratio that the 500-home star/bus systems achieve.

PacBell and US West are citing costs on the order of $1,100 to $1,200 per household to support provision of all services, from POTS (plain old telephone service) to NTSC TV to two-way video. These costs are already down to levels well below the past cost estimates for broadband. If it works, the Ameritech approach would be even less expensive — especially since Ameritech already has fiber running to within two miles of subscriber houses in 60 percent of its territory.

Business applications. As with other telcos, Ameritech’s residential broadband deployment is moving in tandem with deployment for business users. The company expects that within two years the pace of broadband implementation in both market sectors will have reached the point that it will be possible to convert entire network territories over to ATM (asynchronous transfer mode) integrated broadband operations. (For an explanation of ATM, see Vol. 3, No. 7, p. 14.)

According to Mark Farmer, general manager of advanced data services, the company will start with installation of ATM units supplied by AT&T Network Systems in Chicago and at an undisclosed location in Ohio by the second half of next year.

However, relatively few commercial customers are ready to take advantage of ATM just yet. So, Ameritech will deploy services based on SMDS (switched multimegabit digital service) and frame relay technology in 14 cities throughout its five-state territory. “We’re not going out and pre-deploying ATM without firm commitments from customers,” says Farmer.

As the traffic grows, the company will switch over to full ATM systems. The gating item will be a device at the customer premise that converts from ATM to whatever protocol the customer is using internally. Such boxes do not exist yet, but Farmer says he is convinced that they will be available within a year.

Even though the residential and business groups are operating on separate tracks, “if you move through the two timelines (for commercial and residential services), you see a convergence in terms of large-scale deployment by late 1995 or early 1996,” Farmer says. “By then, we’ll be able to run all these services on a common ATM-based platform.”

TRUMPING THE UNIVERSE

With such synergies in the offing, it looks as if the long-standing concerns about whether the nation will get the telecommunications infrastructure it needs when it needs it have been laid to rest. The broadband pipe is coming at warp speed, posing an unforeseen threat to all who have been slow to get on the digital media bandwagon.

But the telco juggernaut raises another concern for anyone who believes competition should be allowed to supplant monopoly control. In light of what is already on the table at the FCC, the question must be asked: Does the explosion in telco activity portend an end to competition before it really begins?

Not if by competition one means use of the network by a variety of gateway providers. The provisions of open network access guarantee there will be competition among service providers with room for an unlimited number of players. But the potential for facilities competition among cable and telephone network operators could well hinge on the extent to which regulators go along with telco interpretations of the existing rules.

Should ongoing cost allocations be pegged to historic narrowband cost structures or to rules that would assign spending in proportions commensurate with actual network applications? Or, put another way, is PacBell right in assuming it can charge telephone rate payers for some 90 percent of its new network costs, just because those costs happen to be below what it would have spent on the old networking approach?

If the cost increment for adding broadband is really marginal in comparison to ongoing capital costs for providing narrowband services, it is hard to see how competitors will stand up to the power of the telcos to deliver video communications at low prices. Thus, regulators’ responses to the arcane questions of new network cost allocations could decide the issue of one wire vs. two long before Congress figures out how to pursue its two-wire agenda.

Fred Dawson, Jonathan Seybold