Future Encounters of the Broadband Kind
More than 1,000 engineers descended on Tampa, Fla. last month to witness the Society of Cable Telecommunications Engineers' annual five-year look at the technical horizon.
The sold-out 2006 Conference on Emerging Technologies covered the technologies that cable might consider friend or foe (wireless and IPTV), as well as those that will help operators free up precious bandwidth.Futurist Carroll: Be more agile
But who better to kick off a show about the future than a futurist?
If part of figuring out the technology future is interpreting consumer shifts, then we should be watching for the "change averse" baby boomers, who just want technology to work.
That's according to keynoter Jim Carroll, the Canadian technology futurist.
"Emerging into purchasing power, and into your customer base, is this generation who thinks differently—is wired differently," Carroll told attendees. And that generation is far more demanding. "They'll expect far more of you, will be constantly pushing you, and will have far less loyalty to you as a brand," he cautioned.
When planning for the future, cable's biggest challenge is "re-skilling," Carroll said. "You know plant—you've been doing it for 20, 30 years. All of a sudden, you're hit with VoIP, codec issues, issues of packet latency—which means it's critical to re-skill the folks who are instrumental in your architectures."
As for the juicy stuff—predictions—Carroll was ready. For starters, with Americans taking 80 billion digital photographs a year, and user-generated video a natural next step, the big numbers will require even more weird prefixes. "We'll be talking about zetabits and yottabits" in 2010, he said.
Plus, consumers will have massive levels of bandwidth, and content choice. That means cable must distinguish itself in customer care, and in QoS. A self-described early adopter when it comes to technology—Carroll's massively networked home contains a 4.5 terabyte server attached to broadband and digital video spigots from Rogers Cable—he urged cable to befriend power users, not penalize them. "The music industry went to war with its customers. Do you really want to go there?"
Bottom line for cable: Agility is the end game. "You must have a plan for short-term longevity—innovate with fluidity, be flexible in design," Carroll said. "Because who knows what can come next?"More storage, more bandwidth
On the age-old question of finding the right balance between storage capacity and bandwidth, one thing came through loud and clear (again) this year: The answer is just "more."
Joe Matarese, CTO of C-COR's Global Strategies division, reported steep price declines in both solid state and disk drive storage—DRAM alone went from $100 per megabyte, in 1991, to less than a dime per megabyte in 2003. Flash memory will follow a similar path; hard disk drive storage is on the steepest slope.
How much storage is ahead? By 2010, watch for the high-end computer to come stocked with 60 empty terabytes of hard disk storage.
It's probably good that all that storage is coming. It looks like we're going to need it. On a worldwide basis, we humans stored more than 5 exabytes of information on print, film, magnetic and optical storage in 2002 alone, according to the University of California/Berkeley's "How Much Information? 2003" study, which Matarese cited.
(Big number refresher: An exabyte is 1,024 petabytes; a petabyte is 1,024 terabytes; and a terabyte is 1,000 Gigabytes.)
And we're going to need big bandwidth even more: According to the same study, 18 exabytes of information moved over electronic channels in 2002—more than three times that which was stored. Keyword: More.The 2010 network
Getting more bandwidth-hungry services into a narrowcast environment is high on the list of Time Warner Cable's interests, and that means moving more deeply into switched digital, noted John Carlucci, the MSO's chief network architect.
"One of the best examples (of increasing available spectrum) is switched digital, which takes advantage of the IP switching fabric, and our QAMs, to only deliver services to nodes in which those services are being used," Carlucci said. "It means we can use our pipes more efficiently."
Carlucci's list of attributes for the cable network in 2010: Diverse, vast, and converged. Capacity improvements, like switched digital, advanced statmuxing, and using H.264 encoding and transcoding, will pave the way for content expansion, and new delivery endpoints, he said.
Adam Tom, president and CEO of RGB Networks, said his view of the cable network in 2010 is one that is fully switched, and outfitted for everything on-demand. Advanced statmuxing and compression will be the norm for high-bandwidth services, as will channel bonding. Just applying advanced statmuxing to on-demand content encoded at a constant bit rate (CBR), Tom said, could yield a 40 percent bandwidth savings.
Getting video over the DOCSIS channel remains a big focus at Cisco Systems, which used ET to create another new acronym on the matter: V-DOC, for "Video DOCSIS."
Harsh Parandekar, a Cisco software development manager, defined "V-DOC" as "video delivery over DOCSIS in the HFC plant—a converged end-to-end IP solution."
Getting there isn't without its challenges, Parandekar and others at ET noted. The cost of DOCSIS ports on existing CMTS gear is high, because operators currently "get" four upstream ports when they want only to add one downstream port. "It isn't as commoditized as QAMs."
Another key problem: Multicast techniques aren't yet supported in gear based on DOCSIS 1.1 and 2.0
Moving to a modular CMTS (M-CMTS) environment will ease the cost burdens, Parandekar said, in the neighborhood of "10 times the bandwidth at one tenth the cost." Similarly, DOCSIS 3.0 contains hooks to resolve the multicast dearth.
Ultimately, Parandekar said, the decision to shift to video-over-DOCSIS depends on your level of belief in IP network proliferation. "If you accept that IP is the emerging protocol of choice for communications, including video, then V-DOC is a practical solution," he concluded.Sizing up telco TV
Finally: Some useable, articulate information about how telco video stacks up, relative to cable video. The goods came from Nimrod Ben-Natan, a VP at Harmonic (which was providing video gear to Verizon, through a relationship with Tellabs—so we're thinking they should know).
So here it is: Three things, unique to telco-delivered video-over-IP.
One: The much-vaunted fast channel change, which, when demonstrated during Ben-Natan's presentation, looked as fast as it used to be, back when the F-connector plugged into the back of the TV set. How fast is fast? Less than 200 milliseconds per channel change.
Two: Aggressive use of tunerless picture-in-picture, where three or more smaller, lower-resolution (200 kbps) video windows appear alongside the main TV channel, to display, say, additional camera angles on a sporting event, or a mosaic of simultaneously airing games.
Three: Targeted advertising. Because the telco network is inherently point-to-point, ads could be spliced into a unicast stream for specific homes, or groups of homes, he said.
Telco video isn't without challenges, though, Ben-Natan said—notably bandwidth. Supporting a home with two HD sets, each eating up 6 Mbps, removes more than half of the available bandwidth on a 20 Mbps connection—the current deployable max for advanced DSL gear.Getting to know the 'residential node'
Back in the RFI days of the NGNA project (circa 2004), there was much discussion of the VNIU, or Video Network Interface Unit. The VNIU, as it was described then, would be affixed to the side of a customer's home to handle the digital-to-analog conversion for every outlet in the home. Among other things, this would eliminate the need for individual set-tops at every coax outlet.
Fast-forward to today, and we're starting to hear more about a similar device, but under a different name: the residential node.
Describing this concept further was WeiMin Zhang, the chief technology officer of BroadLogic Network Technologies. The residential node, he explained, is in essence a gateway device that can be centrally located inside or outside the house. Wherever it ends up, it will create a "translation point" between the coax plant and the in-home coax network. Even if the content being sent down the plant is all or mostly digital, TVs (especially those without digital set-tops) in homes with a residential node will still see those signals as analog. Zhang said such a device will greatly help operators push the transition to digital without leaving analog customers in the dust. And that's not a small group to ignore, as 60 percent to 70 percent of cable homes are still using analog-only services.Waxing wireless
Another hot topic at ET was wireless—as in leveraging wireless to extend the cable plant and to usher in a range of mobile data, voice and video services.
Although WiMAX tends to grab most of the headlines when it comes to high-speed wireless technologies, another, 802.11n, should also be given heavy consideration as operators decide how wireless technology will hook into their traditional HFC networks, said Mike Rude, director of technical marketing at Metalink.
MIMO (multiple input/multiple output) technology is the "cornerstone" of 802.11n, Rude explained. MIMO sends multiple independent signals on the same frequency via different antennas. Similar in approach to DOCSIS 3.0, 802.11n also enlists channel bonding techniques when applied to the 5 GHz spectrum.
802.11n, a wireless platform whose performance roadmap could hit 600 Mbps, is still in the standards process, thanks to squabbling and other delays.
Still, he called on cable operators to consider integrating 802.11n into their own modems, set-tops and multimedia terminal adapters to facilitate the networking of bandwidth-intensive applications such as HDTV. By the same token, the technology will also enable others to bypass the operator and take services directly to the customer. 802.11n "will show up in cable homes with or without an invitation," Rude said. "Make sure your devices have this interface on them."
802.11n is also a technology that will support the development of wireless mesh networks, a subject covered by Bob Scott, director of wireless networking for Scientific-Atlanta, which is approaching the wireless market in partnership with Tropos Networks. Although adding wireless capabilities to the cable plant is the logical next step in the network's evolution, adding wireless for the sake of adding wireless is not how it should be viewed, he said.
"It finally hit us that we were looking at it the wrong way. Consumers don't care about wireless; they care about mobility," Scott said.
Although WiFi, 3G and WiMAX are hot topics, none of these solutions solve all the problems. "There's no one size fits all," Scott explained. "But [wireless mesh] is a great technology for cable," he noted.Traver: Compete for entertainment dollars
As portable devices, podcasts and "over-the-top" services proliferate, the cable industry must do all in its power to maintain and grow its share of the consumer entertainment dollar, said Comcast Media Center SVP and COO Gary Traver, the keynoter for the final day of ET.
He warned, however, that the industry must stay wary of passing technology fads and "separate the hype and reality" before determining which ones will have a long-lasting impact on the bottom line.
"Sometimes in the early stages, we don't know what will be a fad and what will have the sustainable business models. But one of the things that we do know is that on-demand content consumption is at the core."
Traver also was critical of the "triple-play" and "quad-play" verbiage. With service convergence on the horizon, operators should instead think of it as an "integrated single play" that should simplify the way consumers use technologies and services.
"Winning the battle for the media customers of the year 2010 will come down to who does a better job of serving the customer's need for convenience, choice and relevance," he said.