Three Ways to Fail with VoIP: How Not to Engineer Your Networks for IP Telephony Services

It’s every performer’s nightmare: The stage lights go on; the audience applauds politely; and the room goes quiet. The flop sweat burns your eyes; you open your mouth, … and not even the measliest sound squeaks out. In the case of an Internet protocol (IP) telephony provider, the bad dream goes like this: You fire up voice over IP (VoIP) technology on your hybrid fiber/coax (HFC) net, and the subscribers sign up. They pick up their phones … and not even the measliest dialtone results. Or: You fire up VoIP technology with "me too!" glee, and no one signs up. Or: You ramp up in a big way before prepping for the likes of number portability and E911. Subs and the Feds come after you like Freddy Krueger in Nightmare on Elm Street. These are the go-go days of VoIP, and everyone’s going ga-ga over its potential in cable systems. But experts warn that it’s important to remember that as cable develops its IP telephony engineering prowess, it’s fairly easy to fail if you don’t have the right tools and knowledge about "voyp" technology. Way to Fail No. 1: "Me Too" According to those who are digging cable’s frontline VoIP trenches, if you’re not differentiating yourself from the incumbent local exchange carriers (ILECs), that’s the first smooch of death. "We’re pursuing differentiating services, so we don’t want to get into a `me-too’ dial-tone price war with the incumbents," says Cameron Gough, vice president, communications engineering at Comcast. "We’d rather have a value war where it’ll be a very sticky package [with video and data]. So, when someone comes and knocks on the door and offers new video service or new data service, we get people to say, `No, I’m pretty happy with where I am today.’" Comcast is only slowly revealing its VoIP engineering strategy, including three trials planned this year. But Gough makes one thing clear: "What we’re after is a local exchange carrier (LEC) replacement." Cox Communications seems in agreement with the Comcast tactic of refusing to get in a price war with independent LECs, but rather waging the value battle. "I agree with Cameron 100 percent on this," Bill Dame, director of network switch engineering at Cox, says. "His company and my company seem to run parallel lines here, going toward the same end goal, which is an ILEC replacement VoIP." Dame stresses that VoIP technology has proven itself well, but there remain a lot of variables. "In making a choice for VoIP, going through the initial due diligence and picking the best-of-breed components as per CableLabs’ PacketCable specifications, is a wonderful thing," Dame explains. "You can certainly bring a lot of efficiency and economies to bear, but the one thing that folks don’t really realize is that once you go down that road to a best-of-breed architecture, the picking the components, the getting it up originally or initially, is the easy part." Jos� Alegria, general manager of Liberty Cablevision of Puerto Rico, says that his company’s rationale for pursuing PacketCable VoIP instead of session initiation protocol (SIP)-based telephony is that it wants to compete directly with the ILECs in Puerto Rico. "Verizon is the main phone operator, holding 98 percent of the telephony market on the island, so essentially there’s a large market that we can go after. If we can get 12, 15, 20 percent of that market share, certainly we’re going to have a great business that will produce," Alegria says. "We are trying also to bundle our service to compete more effectively with satellite offerings and compete effectively with other ISPs on the island." Liberty Cablevision of Puerto Rico began its VoIP venture in 2001. After seeing the success of Liberty investments in Chile and Argentina with switch-based telephony, it started developing models that resembled switch-based telephony. Late in 2001, it was approached by Net2Phone (see sidebar) that had previous experience with IP telephony. "We had our first meeting in February 2002, and in four months after starting to do the joint venture between our companies, we had our first phone call in June 2002 from Puerto Rico to Denver," Alegria explains. "Right after that, we started doing a test trial, starting with 200 customers, just to test the feasibility of the technical solution, which is basically just doing tones and calls from end to end." Net2Phone provides most of the architectural work, performs the solutions for putting in and installing all the equipment (from soft switches to gateways at Liberty’s headends), and also monitors the operations 24/7. Way to Fail No. 2: Poison the Well There are some nasty ways to ensure that you get a bad rep in the IP telephony biz, and running in with technical blinders on is possibly the best. "We’re finding numerous minefields along the way, which is to be expected in new technology," Cox’s Dame reports. "But as we find the minefields, the minesweepers—our engineers—are out there clearing the way for us. And the future looks very, very good for VoIP and for anyone who really wants to get into it." Dame believes there is sufficient knowledge in the cable industry now that if you have a question, you can certainly find an answer if you talk to the right cable people. Comcast’s Gough also is happy to see cable engineers beefing up their VoIP know-how and stresses the slow, deliberate approach as well: "We want to be very careful not to poison the well and make mistakes early on." Local number portability (LNP) and E911 are all things that people expect of dial-tone, so that’s the premise behind Comcast’s very slow, very measured VoIP tactic this year. "At the end of the year, we’ll take a look and see where we are, and then we may ramp up in a big fashion or continue to learn some more over the next couple of years," Gough promises. "We’ll see how that plays." Way to Fail No. 3: Think Integration Later "The process of system integration is a real challenge, and that could be a formidable obstacle if you want to develop the back-office infrastructure and support and develop the technology management aspects of this," says Joe Jensen, president of Buckeye Telesystems. A unique aspect to the Buckeye story is that it has hawked its telephony wares (from the beginning) toward the commercial side of the business, rather than the traditional cable residential market. In 1998, Buckeye launched a circuit-switched services platform for voice and data. "But in our focus, it was more toward the commercial side of the business. We went after customers who had typically 10 lines or more—typically those that needed data services. "To give you an idea, we have about 950 buildings on our fiber network with that. We have roughly 1,000 accounts, and the monthly revenue average per account is in excess of $1,300," Jensen reports. Liberty’s Alegria agrees with Jensen’s analysis of the integration situation. "There are numerous challenges, especially with the know-how of a local team that has not had previous experience with telephony," Alegria says. "[Having] a hosted turnkey solution, as provided by Net2Phone, has given us the advantage of coming out with a new product—easy to market, easy to install, easy to develop and deploy—as our hosted voice over IP solution." One of the most daunting aspects of integration is interconnection with the public switched telephone network (PSTN). "It’s not just because of the scale or the multiple markets that we’re contemplating," Comcast’s Gough says. "It’s just that the doing of it the first time is not easy, and we found we’re depending heavily on people with experience in the phone world." Another challenging aspect is integration of provisioning, billing, surveillance and monitoring, Gough adds. "If I were in a small operation, I’d really try and find someone that could help me with those aspects," he advises. "Even with our experience in cable telephony, there are a lot of little things that can go wrong that take some experience and a deep understanding of how the phone system in North America works." Cox’s Dame puts it this way: "Telephone has been around for a long time, and it has a lot of regulatory pitfalls and legalities. If experience is not available, it can be overlooked very easily." If you want to do a true ILEC replacement service, expertise is essential, Dame stresses. "You can’t really expect a third party, a vendor company, to come in and hold your hand and go through it," he says. "You’re going to need the help of an experienced consulting engineer or possibly you’ll need to bring the personnel into your company, which we would certainly recommend." Cable is going up against "the 800-pound gorillas of the telephone industry that have had customers for years and years and years," Dame points out. "If you can’t offer really good quality of service, your churn rate’s going to go up high. And chances are, the customers you lose, you won’t get back." If you really want to be an ILEC replacement, then you have a "heck of a long road to go down," Dame warns. "But there is help out there." Laura Hamilton is the editorial director of Communications Technology. Email her at [email protected]. Should You Outsource IP Telephony? Voice over IP telephony is less capital-intensive than circuit-switched telephony. However, IP telephony can prove expensive if it’s not done correctly from Day 1. Operational, engineering, support and project management issues can be daunting to small- and medium-sized ops, so some have taken the outsourcing path. One company in this space is Net2Phone. We spoke to the company’s Sarah Hofstetter, senior VP corporate communications. What kind of integration issues have you addressed with Liberty and the other cable operators? Integration issues vary per cable operator. One of the most important that we have addressed is integrating a VoIP service into the current billing system, (CSG and others). It is important for a provider to be able to offer a unified bill for all of its bundled services, and we have worked on integrating our telephony billing with that of a cable operator’s other services. Another issue we’ve addressed is that of integrating all the various components for a telephony solution. Just because the components are DOCSIS-compliant doesn’t mean that they function correctly when configured as part of a larger telephony solution. When considering managed end-to-end VoIP telephony, what are cable engineers’ biggest concerns? What are the top engineering lessons-learned from the deployment at Liberty Cablevision of Puerto Rico? Many in the industry agree that cable shouldn’t get into a price war with the telcos, but rather a value war. (For example, offering the likes of free voicemail or call waiting, etc., with telephony.) How are the operators Net2Phone is working with addressing this strategy? How does it affect engineering? We feel that providing a host of options to cable operators allows them the freedom to offer the products and services that best suit their customer base. Each operator has to determine what their customers will find the most attractive offer and market it appropriately. Whether it be by price or by value, our job is to make sure that the service, when delivered, is of premium quality. Bottom Line VoIP Pitfalls We feel that providing a host of options to cable operators allows them the freedom to offer the products and services that best suit their customer base. Each operator has to determine what their customers will find the most attractive offer and market it appropriately. Whether it be by price or by value, our job is to make sure that the service, when delivered, is of premium quality. Remember that as cable develops its IP telephony engineering prowess, it’s fairly easy to fail if you don’t have the right tools and knowledge about VoIP technology. From those on the front lines, here are common paths to failure.