Pipeline Profile: Dave Large

SCTE Member Since 1976
Title: Independent Consultant Broadband Background: For three years, Large was co-founder and CTO of triple-play service provider Altrio Communications. He previously served fifteen years as CTO of two different large MSOs. He has twelve years of experience in designing and supervising design of RF equipment, including cable television test equipment and six years of successful consulting to network operators, equipment manufacturers, cities and legal firms in the U.S., Central and South America, Asia and Europe.. What do you consider your biggest success at Altrio? Working with a true "dream team" of talent, we built a system from the ground up and simultaneously launched analog and digital video, telephone, high-speed data and VOD. Within three years we achieved 85 percent digital, 75 percent data and more than 50 percent telephone penetration of connected households, with very low churn rates. What was the biggest lesson learned? Everything that is political, regulatory or financial takes longer than you ever imagined. Is capacity strained today? Why? Yes, for several reasons. First, there is a macro-trend towards viewer-controlled consumption of video. VOD in all its flavors is a part of that. Second, cable operators are adding telephony, a symmetrical service that adds to the already-heavy upstream load. Third, television in the US is evolving towards high definition as the default format and HDTV requires about four times as much capacity as standard definition. Finally, and perhaps most significantly, operators are being required to increase cable modem speeds quickly to stay ahead of competition. With the exception of broadcast HDTV, all of these services require communications to and from individual customers and thus requires capacity increases proportional to the product of penetration and usage rates. As we look to the near-term future, new services such as gaming, video telephony and lots of video and audio streaming from the Internet will continue to drive bandwidth usage. Why do you regard bandwidth requirements for cable modems as particularly significant? Over the past 20-30 years, the RF bandwidth of cable systems has been increasing at an approximately linear rate, while the speed of data services has been increasing at an exponential rate – following the Moore’s law increase in computing and memory capacity. We have gotten away with this initially because of our very high transmission capacity and because of the conversion to HFC architectures that reduced the number of subscribers sharing bandwidth, but the path forward is not as clear. What options do operators have? Several options are under discussion. First, we can increase RF bandwidth once again, though an increase from 860MHz to 1 GHz is a downstream increase of only 17 percent. Second, we can use the bandwidth more efficiently by going from 256 QAM to 1024 QAM, but that offers an increase of only 25 percent of the bandwidth assigned to digital services and requires another 6 dB of C/N, which might be problematic for some systems. Third, we can abandon MEG-2 video compression in favor of MPEG-4, Windows Media or another more efficient algorithm and potentially realize a 2:1 video throughput increase. All three of these options, however, require replacement of all current digital set-top boxes and also are non-compliant with current FCC regulations for one-way digital video services. A fourth option is to convert all analog video to digital and more efficiently use the bandwidth now devoted to basic cable service. This conflicts with current regulations requiring analog carriage of broadcast signals and also requires placing new digital converters on top of every television set we service – several per customer. Switched digital video has gotten a lot of press lately and will certainly be part of the long-term solution. Unfortunately, it requires a two-way subscriber terminal to communicate requests to the headend, and thus is either limited to interactive video services or, alternately, has the same regulatory hurdle as 1024 QAM while requiring the same massive converter rollout as the all-digital option. Subdividing optical nodes in the plant creates no regulatory or premises equipment problems and will also be part of the long-term solution, in my opinion. While it does not increase the instantaneous bandwidth in either direction, it reduces the number of subscribers sharing that bandwidth. Because cable operators devote only a small percentage of downstream bandwidth to interactive services today, node splitting is primarily an upstream solution. Where can operators most efficiently gain the most additional capacity? I think it is time to once again look at the frequencies above 1 GHz, an idea the industry has considered several times in the past. The advantage is that we can add significant new downstream and upstream bandwidth using separate amplifiers for just the new band segments. This takes advantage of existing installed coaxial cables, while avoiding the requirement to upgrade existing active devices to handle even more bandwidth and signal load. Are you happy with the second edition of Modern Cable Television Technology? Writing a book like this is a race against time. The technology of the industry is changing so quickly that the best you can hope for is to capture a point in time and hope to explain it in terms that allow the reader to extrapolate into at least the near future. I hope we did that but, being a bunch of perfectionists, I doubt that any of us are completely satisfied. What was the biggest change in the industry between the two editions? We completely re-wrote the digital video, data and home networking sections; added new chapters on wavelength division multiplexing and FTTx technologies; extensively re-wrote the architecture and telephone sections; and re-worked the subscriber equipment section. In all, about 40 percent of the book is new material. Those changes all reflected areas where the industry has changed most significantly.