Comcast Applies CBR Lessons to VoIP Rollouts

Archives June 2005 Issue Slow and Steady Wins the Race Comcast’s approach to telephony for years has been characterized by caution and conservatism. And that’s no bad thing. By Jim Barthold For years, when asked about their impending telephony rollout, Comcast senior engineers would puff out their chests and do their best Orson Welles impersonation, intoning: “We will sell no voice before its time.” Then Comcast acquired MediaOne and AT&T Broadband and inherited lots of constant bit rate (CBR) telephony, and the engineers’ tune changed slightly. “We will sell no VoIP (voice over Internet protocol) before its time,” they would say, obviously avoiding the fact that Comcast was selling voice services through the old CBR equipment. `Found goods’ and research Never one to waste a revenue opportunity, Comcast continued to operate those widespread CBR systems while developing its packetized telephony technology in its vendors’ labs, its own labs and through the CableLabs PacketCable effort. Now, Comcast is rolling out VoIP services and, surprisingly for a company that maintained it saw no place for CBR in any of its business plans, senior company engineering executives admit they learned a lot from that legacy CBR business. “This (packet telephony) is more complex than our experience from the CBR perspective,” says Charlotte Field, the MSO’s senior vice president of national communications engineering and operations. (Charlotte Field is also this year’s Women in Technology winner.) The complexities were studied in the lab, not in the field, so when it came time to apply packet telephony in a market like Indianapolis, Comcast knew “it’s extremely important to be very proactive with data and respond to those issues that can drive performance,” Field says. Even for Comcast, driven as it was to move forward with a flawless product, the intrinsic difficulties of delivering quality, primary line voice service over IP networks were surprising, says Field. “As Time Warner, Cox (and) ourselves are getting into this business, we’re understanding that what was good enough for data is not good enough for voice,” she says. “We need to focus a lot of energy on the certification of the plant.” Lessons learned The company’s cautious service rollout led to some fast and hard lessons. Voice, as everyone should know by now, is not like data where an impulse hit or burst of packet loss is annoying, but hardly earth-shattering. With voice, there is a need to pre-certify the network before even putting out the first packets. “What we’ve done over the last several years while we’ve been in that CBR business was build a toolset, correlation tools, automatic ticketing systems and the knowledge base,” says Cameron Gough, the MSO’s vice president of communications engineering. “We’re not only leveraging those tools, but we’re also leveraging some of the infrastructure from a people perspective that we grew with the CBR business.” In fact, even though Indianapolis is not and never was a CBR market, the local management center that supports the system on a 24/7 basis is in Detroit, a former MediaOne franchise and a CBR system. “We are also leveraging our Denver-based voice expertise as well as our Denver and Mount Laurel, NJ, IP expertise,” says Gough. Comcast has, of course, taken the IP telephony network experience an extra step. While Field says that the existing hybrid fiber/coax (HFC) architecture was “very satisfactory,” the company found ways to tweak it to improve connectivity. “One of the things that we are doing that’s very different from some of our compatriots is building our own interconnection(s),” she says. “In Indianapolis, we expanded some fiber the opposite way, not down to the households but up into the (public switched telephone) network to interconnect traffic.” A conventional service While the service is packet telephony—Comcast emphasizes the term VoIP can be misleading because their voice packets never touch the public Internet—it doesn’t include all the bells and whistles consumers are being taught to expect from IP telephony. This strategy puts Comcast behind nonfacilities-based VoIP players such as Vonage who offer a variety of different applications to differentiate their VoIP services. The strategy to stick to the knitting and offer conventional phone service is “actually incredibly smart,” says Lindsay Shroth, senior analyst-broadband access, for The Yankee Group. “Even though they have the ability to do SIP (session initiation protocol)-based services, they’ve equipped the solution as an all-you-can-eat cheaper voice service that looks and feels exactly like the phone service that you have today. That is what the mass market consumer wants right this minute.” The typical Vonage customer, she said, is a “tech-savvy risk-taker that doesn’t mind creating the solution. If they pick up the phone and there’s no dial tone, it’s not a big deal. They want the advanced features and functionality now.” They won’t get most of them from Comcast, although the MSO is looking at ways to use its storehouse of content and applications across its IP networks. “As we get further into the future, there are some other things we’ve been thinking about, some sexy new IP stuff,” Gough says. That might include such futuristic—and so far, unrealistic—things as seamless Wi-Fi-cellular roaming or more realistic ringback tones or customized ring tones. For now, the plan is to use the CBR knowledge base to deliver typical phone service, including emergency 911 dialing and battery backup, two items missing on many nonfacilities-based offerings at a lower price than the incumbent local exchange carrier (ILEC). Comcast’s offer does not include self-installation, which is somewhat surprising, considering the MSO’s presence in the retail space with self-install cable modems that could be used as potential bridges to IP phones. “That’s one of the things we’re looking at,” said Field. “But you have to ensure that you have the proper back-office functions (so) that you assign a telephone number in the right rate center with the right emergency services number to go to the right PSAP (public safety answering point) with the right address. Those are things that we do today in our traditional circuit-switched phone service and we are also doing in all of our Comcast digital voice markets, including Indianapolis.” Again, there’s a reliance on the old way of doing things to clear the way for the new services. Because Comcast’s digital phone service doesn’t include the option of taking the phone away from its home base—”our current product is location-specific,” says Gough—self-installation is a secondary priority, more important from a personnel cost-savings than a technological perspective. “We’re currently a little more comfortable sending a truck out with a technician to make sure everything is done correctly and in the right order there,” he says. “There is a proper way to install and an improper way to install,” adds Field, who says the company would prefer to be cautious. “We are looking at doing a couple of trials with friendlies for how difficult it is to do a portion of the self-install.” Taking no chances In other words, Comcast is leaving nothing to chance rolling out its whole-house digital voice offering. It’s not exactly black phone service—a profanity among the engineering ranks at Comcast for years—but it’s also not the high-tech whiz-bang feature set that vendors have been touting. Comcast was predictably silent on which vendors—Cisco, Cedar Point and Motorola are among the announced leaders—are providing what equipment to which market and what features that gear might facilitate. “We built the whole business around our CBR business,” emphasizes Field. “As a result of that, we can get a customer service request to the current provider, issue a local service request, issue a directory listing request, issue a nonpub directory listing request … and launch the E911 to ensure that those records are updated.” So can Verizon or any other ILEC entering the VoIP space, though those players take a backseat to Comcast, says Shroth. “Their offerings right now are still pretty weak,” she said. “Verizon’s product is just something they threw out there to start to test what you could do with a VoIP service … but it’s really not a longstanding product that they’re going to stay in market with.” And even though other cable operators that have been in the VoIP business longer—such as Time Warner and Cablevision—might have more advanced features and certainly have a larger subscriber base, “cable companies don’t compete with each other,” she points out. Shroth gives Comcast’s migratory technology strategy a thumbs up and says frontrunning VoIP providers such as Vonage could do the spadework testing new IP features. “Alternative phone operators can go out and educate people on some of the features and functionality of what VoIP service could potentially provide, and when that gets to market, they’ll start to develop a platform with those features and functionality,” says Shroth. What Comcast is saying is that, despite all the talk in the past, there really were lessons to be learned from CBR, and packetized telephony really is more difficult than anyone could have imagined. Jim Barthold is a contributor to Communications Technology. Reach him at [email protected]. Did this article help you? Email comments to [email protected].