Will the VOIP industry be punished following the Supreme Court's Brand X decision?
That's what some are wondering after the court's classification of cable broadband as an “information service" rather than a "telecommunications service." The worry: The classification may limit what VOIP providers can do if their service is blocked by a broadband provider (see Supremes Sing Cable's Praises).
The last time a VOIP service was blocked by a broadband provider (Madison River), the Federal Communications Commission (FCC) ruled that a "telecommunications service" must remain neutral to all consumer applications flowing over it (see Vonage Victorious in Blocking Case). But not all phone calls are considered part of a "telecommunications service" in the cable world, and they likely won't remain so in the telecom world either.
The Supreme Court in the Brand X decision reaffirmed the FCC’s earlier contention that a broadband “transport” service is indistinguishable from the services (like VOIP) that run over it, and as such must be classified as one thing -- an information service.
The Voice On the Net (VON) Coalition's Staci Pies, for one, thinks this may take away some of the legal leverage a "bring-your-own-access" VOIP provider like Vonage Holdings Corp. might need to effectively seek relief through the FCC.
“The negative here is that it may be easier for the owners of that [broadband] facility -- the ILEC or the cable company -- to discriminate against traffic coming from unaffiliated providers, whether that’s VOIP traffic or some other kind of data,” Pies says. “When there is a problem, there’s very little recourse now to seek an enforcement action.”
Now that many of the same companies that offer broadband service are launching their own vertically integrated VOIP services, the financial incentive to block or tamper with unaffiliated VOIP provider traffic becomes clear (see Does VOIP Business Add Up?).
Vonage remains concerned about VOIP port blocking but plays down the importance of how the courts or the FCC classify broadband service. As such, Vonage believes the Brand X decision will have “no material effect” on its business.
“A lot of people are trying to tie the neutrality issue to telecommunication services law, and we honestly view the [broadband provider] neutrality issue as a separate and distinct issue from what classification your service is,” says Vonage spokeswoman Brooke Schulz.
“The broadband consumer should have a threshold set of assurances that they can rely upon to make sure that their traffic isn’t tampered with,” Schulz says.
But no such set of assurances exists today. While the FCC acted quickly and decisively to stop Madison River from blocking Vonage traffic, the agency has not issued a ruling specifying that broadband providers remain “neutral” and do not discriminate against information services like VOIP.
VOIP service provider Nuvio Corp. last year filed an ex-parte letter with the FCC asking for exactly that sort of thing. In the letter, Nuvio says the FCC should use its Title I authority to prohibit broadband providers from hindering the traffic of unaffiliated VOIP providers.
“Vertically integrated broadband and voice providers have every incentive to discriminate against unaffiliated VOIP providers,” Nuvio CEO Jason Talley states in the letter. “By blocking or degrading access to unaffiliated VOIP services, the vertically integrated firm can create a quality difference in favor of its affiliated voice service.”
“If these vertically integrated firms are allowed to discriminate against unaffiliated VOIP providers, they will almost surely garner the majority share of the VOIP market, and in doing so drive smaller unaffiliated VOIP providers out of the market,” the letter states.
Washington telecommunications attorney Dana Frix says folks may be reading too much into the FCC's intent. “These are really just technicalities,” Frix says.
Frix points out that while the Supreme Court in Brand X calls broadband an information service, it also deferred to the FCC as the best possible authority to rule on such issues.
"The questions the Commission resolved in the order under review involve a “subject matter [that] is technical, complex, and dynamic," Justice Clarence Thomas said in the Brand X decision. “The Commission is in a far better position to address these questions than we are.”
"The FCC could take a look tomorrow and say, 'That thing that we called an information service is in fact going to be called a telecommunications service or is a kangaroo service,' or whatever the hell it wants to call it,” Frix says.
For the past five years, Frix points out, the commission has been very friendly to unaffiliated players like Vonage, consistently enforcing maintenance of the free flow of Internet traffic.
“We’ve got to stay focused on the bigger picture here,” Frix says. “Believe me -- if a cable operator were to block Vonage traffic, the FCC would smack them down immediately.”
If ISDN was not ready for prime time until 1992 in Europe, and it is arguable when it was in the U.S. (your comment about National ISDN) this leaves the impression of a very long development / maturity / cost optimization cycle, and the possibility that many unforseen technologies and trends were in the final mix of ISDN deployments, either in actuality or on the horizon. Even outside of the negative impacts of usage-based pricing models (which I agree with Fred was a major issue), this would seem to complicate the prospects of any technology.
That is just one of many clear takeaways. As a former product manager of an ISDN product, I have other views, but as that will just lead to discussion where there is already clearly expressed disagreement, I won't wander off in that direction; except to suggest that people once again look at all the influences of their perspective including where in the industry structure they sit - it is easy to blame the customer, but the true nature a problem often has many complex tentacles.
Alchemy's post was excellent, so I'll address some other questions.
ch> With respect to modems for dial-up, could you please provide your rationale for why you say ISDN is far better. Do you mean in the sense that a Ferrari is far better than a Taurus, or do you mean in the sense that ISDN has the same cost, paradigm, useability, profile as modems but does it better.
I mean that ISDN is far more efficient. You have a 64000 bps (modulo occasional bit robbing) switching and trunk network. A:D conversion is lossy. So just extend the 64k to the subscriber and it's straightforward. Very low BER, no training time, and 100% efficiency (no A:D losses). It's rare for V.90 to get past 50 kbps on all but the shortest loops, and its upstream is usually about half that. ISDN is symmetric. Plus it's really nice to connect in one second instead of 30. What's to compare? Only carrier pricing, and that's an artifact of the Bells' desire to kill it, not inherent in the technology.
>ch>> It was the RBOCs decision to kill it. <<
ch > Can you please provide an explanation for why they would believe it to be in their best interest to do so.
They wanted ISDN BRI for centrex featurephones. In the 1980s, modems were small potatoes. Come 1993, and suddenly there's a mass market developing for modems for Internet access. Hmmm, ISDN's the best way to do this. But they didn't want this. They had this insane notion that they should treat dialup ISP calls as if they were long distance (switched access, aka Þe Olde Modem Taxe). Now they couldn't stop analog lines from being used for modems. But by 1994 they knew that most non-Centrex ISDN BRIs, especially residential, were being used to dial up ISPs. So they attacked them as a proxy for attacking ISPs. Now any sane person would know that the result would be more analog modem calls to ISPs, not less ISP calls overall, and analog actually cost them more per minute (per many cost studies). But that's how they behaved. Guilt by association.
chucksmaam asks: Also, could you provide an opinion about when you believe ISDN as a technology was ready for prime time.
I'm not FGoldstein but I fought the ISDN wars for a decade. ISDN as a technology was ready for prime time in Europe in 1992. That's when the ETSI Euro-ISDN specs were released, the labs were set up to test CPE gear and assign telecom CE Marks, and when Alcatel, Siemens, and Ericsson had finished implementing to it. There were strange and wonderful country-specific variants before that. Germany rolled early with 1TR6 which looked more like the CCITT 1984 Red Book spec. The UK had some really strange variants like DASS2 that were even less Q.931-like.
In the US, you could get AT&T Custom and Nortel Custom Basic Rate in some areas in the early 1990's. You could argue that National ISDN for BRI was so screwed up that it was never ready for prime time. Service Profile IDs and a very lousy Nortel implementation made self-installation really painful. On Primary Rate, AT&T had a really bomb-proof #4ESS implementation from Pub 41449/41459 in 1989. Siemens was the only National ISDN-2 implementation on Primary Rate for a long time and there are only a hand-full of Siemens switches deployed.
>> it's far better than analog POTS for voice lines, and far superior to modems for dialup. <<
With respect to modems for dial-up, could you please provide your rationale for why you say ISDN is far better. Do you mean in the sense that a Ferrari is far better than a Taurus, or do you mean in the sense that ISDN has the same cost, paradigm, useability, profile as modems but does it better.
>> It was the RBOCs decision to kill it. <<
Can you please provide an explanation for why they would believe it to be in their best interest to do so.
Also, could you provide an opinion about when you believe ISDN as a technology was ready for prime time.
mg> "ISDN was introduced in 1985, when a state of the art modem was 1,200 bits per second, so 128,000 bits per second (2 x 64kb/s B channels bonded) was not an order of magnitude improvement; it was two orders of magnitude improvement."
mg> If this is even half true, it shows the power of competition around open specs to grow versus closed, monopolistic, specs. In time, all monopolies will get bypassed by superior collective progress.
The original statement was not quite true. In 1985, 2400 bps dialup modems were commonplace. The V.32 standard for 9600 bps dialup modems was written around 1984, but products were not widely available until a couple of years later. 2400 bps was the fastest speed that a modem could use over dialup without active equalization. That was costly, given the chip technology of the day. So it took another iteration of Moore's law or so, another round of chip technology, to make V.32 affordable. V.34, with 28.8 kbps, came in the 1990s, after ISDN was out. V.90 was actually inspired, I think, by the "data over voicer bearer" ISDN hack that I and others pushed as a workaround for demented tariffs in NYNEX and GTElands.
ISDN was not suddenly "introduced" at any point. There were some early trials in Europe around 1985-1986, though the physical layers were still firming up. US trials really got under way around 1987.
ISDN was, of course, an international standards project, as open as anything in the telecom world. So were the V.-series modems. What hurt in the USA, though not Europe, was the vendors' insistence on proprietary, vs. standards-based, variants. The ITU standards, even the "mature" 1998 Blue Book, necessarily had room for national variations. Things like telephone number formats vary by country, not to mention rules concerning things like privacy, and backwards-compatibility issues with the PSTN. So ETSI went off and wrote a Euro-ISDN standard, which was widely adopted across the EU and elsewhere that used European telecom standards.
In the US at the time, there were seven RBOCs and only two big CO vendors, AT&T and Nortel. That let the vendors call the shots. Both CO vendors had built pre-standard "proprietary" ISDN variants, differing mainly in some details of the D-channel signaling protocols, especially around supplementary services. (Nortel also treated the two B channels as separate Terminal Endpoints, a cheap and ugly way to recycle P-set code.) This got feature-rich sets onto the market before Q.932 could be finished. But it meant that early ISDN TE makers had to support both "5ESS" and "Nortel" proprietary code. The 5E version was frankly a lot nicer, as well as more respectful of the CCITT architecture.
Bellcore tried to fix this with "National ISDN". But Nortel hijacked that process and NI-1 ended up looking a lot like Nortel Proprietary. So come 1994, with ISDN deployed fairly widely, TE vendors had to support three variants. Nortel proprietary was fading into NI-1, but AT&T proprietary remained in place because it was so much better. And none of it was really well documented: To make workable TE, you had to know the folklore. Americans (whose engineers are mostly lousy writers) understood that, but it drove Europeans batty-bats, because their (Germanic?) custom was for everyone to follow rigid written specs.
So the standards process was open, and it worked in Europe, but vendors who found it inconvenient were able to subvert it in the United States. Much as it's easy to blame the Bells for screwing things up -- and they screwed up the rollout of ISDN miserably, in what I think was deliberate sabotage -- the lack of standardization was really AT&T and Nortel's fault. Still, the lack of good standardization didn't kill the BRI (the PRI is thriving) in the States, since vendors learned to deal with it. Nor was it inferior technology -- it's far better than analog POTS for voice lines, and far superior to modems for dialup. It was the RBOCs decision to kill it.
"ISDN was introduced in 1985, when a state of the art modem was 1,200 bits per second, so 128,000 bits per second (2 x 64kb/s B channels bonded) was not an order of magnitude improvement; it was two orders of magnitude improvement."
If this is even half true, it shows the power of competition around open specs to grow versus closed, monopolistic, specs. In time, all monopolies will get bypassed by superior collective progress.
HeavyDuty writes: ISDN was introduced in 1985, when a state of the art modem was 1,200 bits per second, so 128,000 bits per second (2 x 64kb/s B channels bonded) was not an order of magnitude improvement; it was two orders of magnitude improvement.
This doesn't really map onto the true history of ISDN very well. In 1985, the physical layer of BRI wasn't standardized. The CCITT Red Book Q.931 from 1984 did not look at all like the Blue Book Q.931 from 1988. Anyone who deployed "ISDN" in the 1980's did strange and wonderful homebrews. ETSI was the first truly interoperable residential ISDN variant and that really didn't happen until 1992.
This is like seeing dinosaurs in a Raquel Welch caveman movie. "Bonded" B channel technology also didn't exist in 1985. The BONDing group didn't have a viable spec until 1993. Unless you rolled your own protocol or used X.25, there really wasn't a way to aggregate 2 bearer channels that didn't maintain time integrity across the PSTN until then.
Mosaic wasn't written until 1993 so demand for residential internet services (data) wasn't going to happen until the mid-1990's. By then, cheap modems had advanced to the point (V.32 @ 9600) where a 64K BRI line didn't have significant advantage over a dialup modem line. By 1999 when the internet became popular, with V.90 modem technology given away for free in every PC there was virtually no advantage.
If you simply track standards activities I believe you will arrive at a different answer than me on this issue. My understanding is that the first 9600 bps modem was demonstrated in 1971, and I know for a fact (unless I am loosing my mind) that I was using Codex 2660/2680s in the 1984 time frame.
>> , but at a cost that made it prohibitive for all but the leased line services that used it. <<
Well yes, this is correct.
ISDN was probably cheaper than leased lines, but was it affordable by the average PC user? Even in 1998 some RBOCs were still offering usage-based ISDN, not really a very accessible technology from a cost perspective.
So we need to start comparing apples to apples. Two examples would be:
-When a technology is standards-based and interoperable solutions exist from multiple vendors; the performance over unconditioned lines; and cost.
Another example would be: -Non standards technology; performance over conditioned lines; and cost.
I consider the performance over non-conditioned lines to be particularly important to this conversation. I seem to recall that when my ISDN line was installed in the late 90's, apart from the estimated 100 man hours of installation time required in passing papers around internally (as told to me by the installer) the line had to be conditioned as well - a cost that I had to bare one way or the other.
So if we are going to assert that one technology is more expensive than another, let's examine all the costs and who the technology is really accessible to, and then we can start discussing from a common baseline what the alternatives were to various target groups.
>> "seem to recall that ISDN was quite different from vendor to vendor..."
That was intentional obsfucation by the RBOCs, because in Europe and California, where.........<<
If that is true, then that is simply an argument for the power/value of overlay services/technologues, which of course the common modem is one example of; perhaps it is a an argument for a strong PUC as well. However, given that ISDN handsets working entirely within a local in-building telephone network were not interoperable among different PBX vendors, I am skeptical about whether the RBOCs are entirely to blame on this issue. I am not familiar with what year the common standards were implemented as a result of the California PUC action, but I would be very interested to know. I am guessing it was later than 1984, but if wrong, would appreciate knowing.
7> Rate of return regulation is long gone. To return to it would be really, really, really painful. In particular, it would be painful for the RBOC shareholders. They bought on a Rate Cap basis and the regulation would wipe out their growth.
I do not propose putting the RBOC stockholders under Rate of Return. I have proposed having the ILECs spin off their outside plant, along with most of their debt, into separate corporations (LoopCos). These would be rate-of-return regulated, and just the thing for widows and orphans to invest in. (Low-risk stocks are growing scarce.) The corporate residuaries (ServiceCos) would be freed up from some of their existing regulatory obligatons, would not be subject to rate of return, and could concentrate on making more profits for their debt-reduced shares.
Win/Win. Especially for shareholders, who would get a one-time choice of which shares they wanted (as happened in the AT&T divestiture), to match their risk/reward goals.
"19.2K modems were available at least as early as 1984..."
A 9.6kb/s modem was available, but at a cost that made it prohibitive for all but the leased line services that used it.
"seem to recall that ISDN was quite different from vendor to vendor..."
That was intentional obsfucation by the RBOCs, because in Europe and California, where the PUC forced Pacific Bell (before SBC's purchase of PacBell & the CA legislature; reined in the PUC) to make BRI available, common standards were adopted causing ISDN to work as well as POTS.
The blogs and comments are the opinions only of the writers and do not reflect the views of Light Reading. They are no substitute for your own research and should not be relied upon for trading or any other purpose.
To save this item to your list of favorite Light Reading content so you can find it later in your Profile page, click the "Save It" button next to the item.
If you found this interesting or useful, please use the links to the services below to share it with other readers. You will need a free account with each service to share an item via that service.
Digg
Del.icio.us
Reddit
