Cable telephony - parasite or profit?
Fred Goldstein, March 2004
... The action is moving instead to parasitic applications of VoIP, the best-known of which is Vonage...
.
...They have no local loop costs, because the subscriber is paying the entire cost of the underlying broadband service...
...Cable companies can allow third-party CLECs to rent access to their networks in order to provide service over PacketCable, in effect providing wholesale local loops...
...cable companies should voluntarily open up their networks to competing ISPs, to provide additional sales channels they can use against the ILECs...
The options open to competitive carriers continue to grow narrower, as the Powell FCC and he DC Circuit Court conspire to shut out Competitive Local Exchange Carrier access to the Unbundled Network Elements that the Telecom Act ostensibly provided for. Powell, however, seems to think the the existence of VoIP is enough competition, since it theoretically can be used over both ILEC-provided DSL and cable modems, the currently-blessed duopoly. CLECs are left with few options, unless they are fortunate enough to be able to become cable overbuilders, or have a protective state commission that independently enforces the law. While there are still avenues by which a CLEC can make money using some unbundled network elements, the FCC and DC Circuit have conspired to create enough regulatory uncertainty to put almost anyone's business plan in doubt.
Cable companies are the other winners for now, very well positioned to compete with incumbent LECs in both voice and data markets, but most have been slow to do so. Cable modems, to be sure, are practically standard equipment. But cable telephony remains surprisingly elusive. Perhaps the cable companies have been waiting for the technology to mature. In the late 1990s, TDM-based dedicated cable telephony systems (such as Tellabs 2300 and Arris' Cornerstone) became available. Using telco-standard GR-303 attachment to a central office switch, they provide high-quality dialtone service. Several million lines have been deployed, but they are no longer in fashion for new installations.
PacketCable vs. parasitic VoIP
A newer standard, PacketCable, has been adopted by Cable Labs and provides for high-quality telephone service across a cable modem infrastructure. PacketCable uses VoIP, with prioritized bandwidth (a feature of DOCSIS 1.1, a now-common cable modem standard) to ensure call quality. Because it is a competitively-produced standard for high-volume products, PacketCable is very inexpensive to deploy. It adds perhaps $50 to the price of a cable modem; it needs no head end terminal of its own, simply a compatible VoIP telephone switch to attach to the Cable Modem Termination System. (Other telephone switches can use it through a gateway. PacketCable is based on MGCP, a fairly popular VoIP standard.)
But that's not where the excitement is. The action is moving instead to parasitic applications of VoIP, the best-known provider of which is Vonage. For a company with only a modest number of subscribers (about 100,000 as of this writing), Vonage is scaring the industry like a mouse in the elephant cage. Not only does it have other startup competitors, but big companies such as AT&T and Qwest are also planning to get in on the act. Parasitic VoIP runs over DSL or cable. It doesn't take advantage of PacketCable's prioritization, so its quality is not guaranteed. It doesn't connect to the CMTS at the head end; it usually requires some amount of public Internet connectivity. It's not installed by the cable company, or even with the knowledge or consent of the cable company. It's just an application running over the braodband service that the subscriber is already paying for. It's parasitic in a technical sense because it feeds off of bandwidth that the subscriber has already purchased, rather than having its own.
From a customer perspective, PacketCable offers a higher-quality service, capable of meeting "lifeline" standards. "Embedded Multimedia Terminal Adapters", as phone-equipped cable modems are formally known, are even available with built-in batteries, substituting for the line-powered service used by traditional telephony as well as most of the proprietary cable telephony systems. Services like E911 work perfectly well, since the cable operator is able to provision the service with the necessary information. PacketCable telephony service is typically offered under tariff, with the cable operator having CLEC status.
Parasitic services today operate as "information" services, not requiring CLEC certification; they typically work with CLEC wholesalers to provide PSTN connectivity and telephone numbers. Their service quality is more erratic. They are more likely to use compressed voice, and because they do not get priority, their voice packets can be discarded when congestion occurs anywhere between the subscriber and the PSTN gateway. At best they sound almost as good as Plain Old Telephone Service; at worst they sound like "bad cellular". Connection quality can change within the course of a call. Fax support is sometimes built in; other modems however are unlikely to work well at all. The main attraction is price: They have no local loop costs, because the subscriber is paying the entire cost of the underlying broadband service. They are currently able to avoid paying the switched access charges that long-distance carriers pay to the LECs at both ends of a call, because they take advantage of the "exemption" that was created to allow dial-up Internet (and other information service) access to be calls to be placed as local. (This is on the table again in the FCC's pending VoIP rulemaking.) So they usually feature either unlimited-use or large block-of-time domestic long distance plans.
Cable operators can team up to compete
Cable operators should look at PacketCable as an opportunity. Many are not prepared to go into the telephone business on their own; it's not their specialty. Legally, they are not required to open up their networks to competitive telecom operators or ISPs. But from a business perspective, they should! Cable companies can allow third-party CLECs to rent access to their networks in order to provide service over PacketCable, in effect providing wholesale local loops. The CLECs will then be responsible for hauling and switching the calls from the CMTS back to the PSTN, making the necessary arrangements to interface to ILECs, handing the customer support calls, installing the eMTAs at customer premises, and doing all of the other heavy lifting. For this, a CLEC should be willing to pay a decent fraction of the ILEC unbundled loop rate! It's a win-win situation. The CLEC gets bypass the ILEC en route to the subscriber, and gets a high-quality connection. The cable company gets incremental revenue, say $5-10/month per CLEC telephone subscriber. It's less than they'd get if they provided the full telephone service, but it's a lot more than they get from Vonage or other parasitic operators (i.e., zero).
CLECs, of course, are often affiliated with ISPs, and would often like to sell data services as well. As I've noted before (including in a 1999 Multichannel News opinion piece that unfortunately is no longer on line), cable companies should voluntarily open up their networks to competing ISPs, to provide additional sales channels they can use against the ILECs, and to reduce the incentive to overbuild. So an ISP-CLEC might want to provide the cable modem and PacketCable service together, paying the underlying cable company for the privilege.
There are some minor technical details that would need to be worked out in each case, but there is already ample precedent for multi-ISP cable modem networks. Multiple CLECs on a cable plant should be even easier. This is a win-win situation for everyone except the ILECs.
