Building for cable’s future
We are proud to present the seventh edition of CED’s Broadband 50: the fifty most prominent companies, technologies, trends and people who defined, shaped and epitomized the broadband industry in 2007, or threaten to do so in 2008.
A long list of potential candidates is compiled by CED scribes and advisors, who discuss, argue over, weigh, bandy about, dispute, and hash over the list.
CED’s staff, supplemented by a secret panel of industry luminaries, then assesses and evaluates every candidate, voting on the relative importance of each. The votes are compiled and tallied, and the candidates are then ranked in order. Many simply do not make the cut. It’s a process.
1: Switched digital video saves the day!
Faster than a speeding bullet! Able to leap over large bandwidth constraints! It’s switched digital video (SDV) to the rescue for cable operators.
The first SDV trial – by Time Warner Cable in Austin, Texas – was way back in 2004, so the technology isn’t exactly brand-spanking new. But, like digital simulcast a few years ago, the technology seems to have reached the tipping point this year for cable operators who want to save precious bandwidth.
With SDV, cable operators can send just the digital signals being watched in a service group or node, instead of the entire lineup of channels. The reclaimed bandwidth goes toward HD and other digital services.
Cox Communications jumped on the SDV bandwagon this year with its first deployment using technology from BigBand in its Northern Virginia system. In November, Charter Communications announced an SDV trial in Los Angeles.
So all of the top five MSOs are in various stages with SDV. Cablevision announced the single largest SDV deployment to date earlier this year, also with BigBand, when it installed the technology across its New York City metropolitan area footprint.
Time Warner Cable is aggressively rolling out SDV since that first trial in Austin. The company’s stated goal is to have SDV in half of its divisions by the end of this year.
While Comcast has lagged a little in terms of SDV, it picked up steam in 2007 with two trials in its Denver and New Jersey systems. Those tests use Arris and Harmonic’s edgeQAMs to deliver SDV, VOD and related services over a shared infrastructure.
Cox, which is using BigBand’s third-generation edgeQAM, said it will start looking at sharing resources between SDV and VOD via edgeQAMs next year.
While other vendors are in the switched digital video space, BigBand has completed or is in the process of deploying SDV with five operators across 20 cable systems that collectively pass more than 11 million households. But this is no longer just BigBand’s game. Watch for Motorola and Scientific Atlanta to pick up steam, too. – Mike Robuck
2: ‘The bandwidth dance’
Look, there’s never enough bandwidth, but it seems like there’s always enough to get by. So far.
In the old days, increasing bandwidth meant upping the amplification – 330 MHz to 450 MHz, then 450 MHz to 550 MHz, then 550 MHz to 750 MHz. Maybe even some 860 MHz, and, if you’re Cox, some 1 GHz. The upper spectral boundary went up as more and more basic and pay channels joined the lineup.
But that was before HD.
Today’s competitive smack-talk for HD goes like this (over and over and over): 100 channels. Spectrally, using existing (MPEG-2) compression, that’s fifty 6-MHz channels (assuming two streams compressed to 19.2 Mbps). OK, 750 minus 550 divided by 6…oops: thirty-three 6-MHz channels in the digital domain, or 17 short of what it’d take to do 100 channels of HD, right now.
Enter “the tools in the toolbox,” which was pretty much the bandwidth story of 2005-2006: Split nodes. Statmux more aggressively. Take back analog channels. Go to 1 GHz. Use a digital video switch.
This year, the demand side of the bandwidth story got HD-crazier, and the supply side picked up a few more tools. Well, technically, one tool was sharpened, and an expected tool started to roll out (see #4).
The tool that was sharpened, at least for one MSO, was a largely overlooked member of the bandwidth preservation family: The MPEG-2 encoder. Yes, MPEG-2. Turns out there’s still some juice in that wringer, at least for the Comcast Media Center (see #46).
At press time, Comcast was actively experimenting with an unnamed encoder maker – can’t imagine who! – that aims to improve the bandwidth efficiency of MPEG-2 by 30 percent. That’s good, right now, while HDTV is still largely a volume game. But our spies say it’ll be even better in the next chapter of HD – picture quality. – Leslie Ellis
3: HD: More, more, more...
Nearly two decades after the modern concept of HDTV was floated, HD became hot in the consumer market in 2007. HD TV sets are selling like hotcakes. Depending on who’s providing the estimate (and estimates vary wildly), there are as many as 30 million HD sets in the U.S., and household penetration hovers near 25 percent.
The direct broadcast satellite (DBS) companies have been baiting the market with ads about all-digital and HD. DirecTV called 2007 the “year of HD” and vowed to have 100 HD channels before the year was out.
That provocation got a big response from cable. In June, Cox President Pat Esser said he’d challenged his engineering team to make 50 channels of HD available by the end of this year, and 100 channels by the end of next. Every major MSO is working toward expanding its HD lineups.
Then Verizon recently upped the ante by setting a goal of 150 HD channels on FiOS TV by the end of 2008.
The fly in the ointment: While HD may encourage migration to the digital tier, it really isn’t monetizable. Its value is all defensive – not losing subscribers with expensive new TV sets dazzled by the promise of a richer HD lineup offered by someone else.
Most cable companies were well on their way toward their year-end 2007 goals of 30, 40, or 50 HD channels, having found ways to scratch out the bandwidth (see #2).
Comcast turned the whole conversation into farce by vowing to have 800 channels of HD by the end of 2008. It arrived at that number by adding in all the on-demand HD content it’s amassing. But Comcast raises a legitimate question: in the age of SDV, what’s a “channel” anyway? So that’ll kick up some competitive heat, if only to challenge the math. –Brian Santo
4: MPEG-4: HBO’s inauguration
The shock value has long-since worn off, but it’s still one of the bigger video-geek stories of the year: HBO’s June bombshell, dropped during a breakfast panel at the SCTE Cable-Tec Expo, that it’ll transmit its upcoming 26-channel HDTV lineup exclusively in MPEG-4. As in, not in MPEG-2, the compression format known by the 50 million digital set-tops inside cable and satellite homes right now.
Somebody had to go first, if only for sheer economic necessity. Especially fitting that it was HBO, first at satellite transmission, in 1975, and first to put up an HD feed.
Transmitting its 26-channel HD lineup in MPEG-2 would’ve required HBO to find space on six or seven additional satellite transponders – double what it currently uses. Satellite transponders go for about $125,000 per month.
Once those MPEG-4-compressed streams get on the ground, though, they’re gonna need some work. Otherwise, they’re unusable by the aforementioned giant installed base of digital boxes that don’t know an MPEG-4 stream from a pumpkin.
Enter the “transcoder,” which is really more of a decoder (from MPEG-4 back to digital baseband) followed by an encoder (to re-code into MPEG-2). Piles of companies showed them at this year’s National Association of Broadcasters Show in April. As technologies go, MPEG-4 encoders are the new black (they and edgeQAMs; see #12).
HBO’s June shocker also included its stance on compression rate: 8 Mbps for HDTV (7 Mbps payload, 1 Mbps overhead). Comparatively, that’s conservative – AT&T is looking for a good 6 Mbps squeeze; vendors at NAB showed an HD stream squished as low as 4 Mbps.
By late summer, HBO quietly announced its choice of Motorola for the gear it and its customers need to receive MPEG-4 encoded and encrypted streams. It’s a pretty safe bet that more and more programmers will start announcing plans to go to MPEG-4 next year. –LE
5: Digital transition: Thank you, sir, may I have another...
Remember when you were a teenager, and your punishment for something you didn’t think you did in the first place was a three-day grounding – but you were half-glad about it, because it wasn’t three weeks? And – whew – you were still allowed to go on the class trip?
That’s the kind of sentiment underlying the FCC’s September decision to “let” cable operators dual-carry broadcast signals in analog and only one digital format (SD or HD) for three years after the 2009 digital transition.
In a world bulging with 100-channel promises of fat, plush HD channels, it doesn’t seem like much of a Hallelujah to be instructed to carry something in essentially three ways, for three years. It’s already job-number-one for most chief tech people to keep the plant sustainable for more HD, let alone channel bonding, and every other new service glimmer.
The Hallelujah is that the three-year passage didn’t say “perpetually,” as in forever, which had been a possible outcome. And it didn’t stipulate that operators had to carry “all content bits,” which would’ve meant skirmishes when considering things like statmuxing and re-encoding. So…um…thanks? – LE
6: Kevin Martin: Doesn’t hate cable. Really.
FCC Chairman Kevin Martin has been like Wile E. Coyote in one of those old Road Runner cartoons: Behind every rock, or standing on every overhang, Martin has been there with yet another wacky anti-cable plan.
Martin either has large cojones or a severe case of hubris, because when he appeared before the cable industry at The Cable Show ’07 this year, he proclaimed that he really doesn’t have it in for cable.
Martin’s constant tilts at the cable industry – whether it was dual carriage, a 30 percent cap on the number of Comcast subscribers, or his theory that cable networks showed too many junk food ads to kids – would almost be amusing if it weren’t for his influential position and the fact that the regulatory battles have been costly, in terms of both time and money, for the industry and the NCTA.
Martin has developed almost a fetish over a la carte pricing, despite evidence that paying for channels individually would actually cost more than the current set up of having the more popular channels carry the less popular ones.
In September, the FCC voted unanimously to require cable operators to carry local TV stations in both analog and digital formats for three years after the analog broadcasts cease on Feb. 17, 2009. Martin initially advocated for perpetual dual carriage without the three-year sunset, (see #5).
And to end the year on a bang, Martin dove deep into the Cable Communications Act of 1984 and brought up a provision that gives the FCC broad regulatory authority over cable if the industry becomes available to 70 percent of households, and 70 percent of those homes subscribe to cable.
He claims cable has exceeded both marks. NCTA President Kyle McSlarrow insisted cable serves well less than 70 percent of the nation’s households.
Rest assured that in the final days of the Bush administration, the wily Martin will have a few more tricks up his sleeve for the cable industry. –MR
7: Google: Like Kudzu
Google extended its reign as Internet King this year, with a few surprises and a well-known setback tacked on to its dynasty.
In November, the company’s $3.1 billion bid to acquire DoubleClick was put on hold when the European Commission refused to approve the deal on competitive grounds. Opposition to the potential deal in the U.S. focuses more on the privacy aspect of it – how much information Google can know about us and who the company can share that data with.
Then, Google roiled the mobile market by announcing its interest in buying spectrum and then introducing the Android handset. It’s preparing to bid for a slice of the 700 MHz band in the FCC’s wireless spectrum auction in January. In July, Google urged the Commission to adopt four open platforms as part of the license conditions: open applications, devices, services and networks. The FCC has decided to support two of the open access principles: applications and devices.
Over-the-air broadcasters are wary of Google’s interest in unused spectrum (within the frequencies of TV channels 2 to 51) that will likely be auctioned off by the FCC in 2009, after the digital transition. Google, as well as other Internet and wireless bigwigs, wants a chunk of broadcast’s VHF and UHF spectrum because the frequencies are also suitable for transmitting video and data wirelessly to mobile phones, PCs and TVs.
As for Android, it’s a collaborative development of Google, T-Mobile, HTC, Qualcomm, Motorola and other companies through the Open Handset Alliance. Google describes the Android platform as a mobile “software stack” that consists of an operating system, middleware, interface and apps. The first Android cell phones are expected to hit stores in the second half of 2008.
On the advertising side, Google established a strategic relationship with Nielsen to bring demographic data to Google TV Ads and introduced Gadget Ads, an interactive ad format. And in April, EchoStar and Google partnered up to introduce the first automated system for buying, selling, delivering and measuring TV ads on the satellite provider’s 125 national programming networks. –Traci Patterson
8: Business services: Flocking to MEF
The Metro Ethernet Forum (MEF) was once an exclusive club for the telco operators, but that started to change last year when Cablevision’s Optimum Lightpath became the first cable operation to join the worldwide forum.
In December of last year, Arris became the 100th member of MEF, and CableLabs, Comcast, SuddenLink and Brighthouse Networks joined this past summer. MEF provides certification waves for both service provider networks and vendors’ products.
Referring to the MEF, Kristine Faulkner, vice president of product development and management for Cox Business Services, said, “It’s about providing assurances for customers that the Ethernet services we’re delivering, and the equipment we’re delivering, are compliant and attain a certain level of performance that our customers expect.”
Both telcos and cable operators are rushing in to provide Carrier Ethernet services to financial institutions, banks and other business entities. According to Infonetics Research, worldwide Ethernet services grew to $5.9 billion in 2005, and it’s expected to increase another 280 percent from 2005 to 2009 to $22.5 billion. Infonetics Research also said Carrier Ethernet switch and router manufacturer revenues grew 132 percent from 2005 to 2006 with a forecast of $9.1 billion by 2010.
Carrier Ethernet offers service providers flexible, standardized carrier-class services such as point-to-point E-Line services, which include Ethernet Private Lines, Virtual Private Lines and Ethernet Internet Access, and multipoint E-LAN services. E-LAN services include MultiPoint Layer2 VPNs and transparent LANs.
This past summer, cable operators were able to cross the MEF 14 certification line in a dead heat with some of their competitors. RCN, Optimum Lightpath and Time Warner Cable earned their MEF 14 stripes alongside AT&T, Verizon and Embarq. Lastly, Optimum Lightpath won MEF’s 2007 Regional Carrier Ethernet Service Provider of the Year award over the summer. – MR
9: DOCSIS 3.0: Patience please
To give you an idea of how big a deal buying more bandwidth is, this is an adjunct to the bandwidth issue, also in the top 10. DOCSIS 3.0 will increase bandwidth on the downstream (to 160 Mbps, at a minimum) and especially on the highly constrained upstream (to 120 Mbps), which will be critical as it becomes increasingly more common for subscribers to upload video.
The speed enhancements are derived largely from channel bonding. DOCSIS 3.0 will also have provisions for IPv6, and modular CMTS technology.
The development of DOCSIS 3.0 technology is supposedly on an accelerated schedule. Until a few months ago, the expectation was that the first deployments might occur in 2009, but now the first commercial rollouts may happen in 2008.
Implementing DOCSIS 3.0 is a priority, especially for cable operators who are competing with Verizon. Verizon is offering a tier at 15 Mbps upstream and 2 Mbps downstream for $49.95, which is faster than most cable operators volunteer to offer without competition, but can still match if they want to. But then Verizon is offering another at 30 Mbps/5 Mbps, seemingly just to twit cable, because at $179.95/month, Verizon can’t be expecting all that many takers.
Who needs 30 Mbps? Who knows? But it’s good to be the fastest, because when you’re the fastest, you get to say you’re the fastest. Avis got famous claiming to be number-two, but bragging about being the second fastest broadband provider is just lame.
Numerous companies – Cablevision, Comcast, Starhub (Singapore), Multikabel (The Netherlands), Videotron (Canada) among them – are experimenting, trialing, testing various aspects of DOCSIS 3.0 technology. Comcast and Virgin Media are among those on the record with plans to roll out the technology in 2008.
CableLabs began Wave 56 of testing in October; results should be available in a few days (the first week in December), but few expect any qualifications of CMTS equipment. CableLabs and the vendors involved were planning to use the session to identify interoperability concerns and to figure out what engineers need to tweak on their respective systems.
That creates an expectation that there might be qualifications as early as Wave 57, the results of which won’t be announced until the middle of March.
A commercial rollout is unlikely before the second quarter of 2008, then. And even if all goes well in Wave 57, which is far from guaranteed, and some operators jump on the opportunity, deployment of DOCSIS 3.0 technology is unlikely to become widespread until 2009. That’s the accelerated schedule.
By the way, broadband in a dozen other countries is way faster for way less money. We’re just saying. –BRS
10: Telco TV: VoIP, huh? Boom! In your face!
In the third quarter, competition from the telcos started affecting the cable industry’s video and broadband subscriber statistics – and their stock prices.
Comcast announced that competition from Verizon’s FiOS and AT&T’s U-verse began biting into subscriber growth in its Q3, and the stock market reacted by dinging all U.S. cable stocks, by anywhere from 5 percent to over 20 percent. Somebody’s taking this competition stuff seriously.
Verizon continues to chug along with FiOS, pretty much on plan. The company has 1.3 million FiOS TV customers (and another 200,000 or so satellite customers) as of the end of the third quarter.
AT&T reported 126,000 customers subscribing to its U-verse video services by the end of Q3, up 51,000 in just three months. AT&T has been expanding rapidly, and the company said that puts to rest any concerns about whether U-verse can scale. But the company also experienced a day-long service failure throughout its entire footprint that was never adequately explained.
AT&T also had to acknowledge that upgrading systems in the acquired BellSouth territory will cost more than expected, which will in turn slow the U-verse rollout a bit, so that AT&T will have 17 million homes passed by the end of 2008, instead of 18 million.
And of course, dozens and dozens of smaller phone companies are offering video.
The interesting thing is that the telcos are pricing their video products roughly the same as cable. The whole point of competition was that it was supposed to force prices down, wasn’t it? –BRS
11: OpenCable Platform: Good-bye OCAP!
As industry-sanctioned nomenclature goes, OCAP is out. At a keynote session at this year’s SCTE Cable-Tec Expo, NCTA Chairman Kyle McSlarrow put the big kibosh on the four-letter acronym: “I have, tongue-in-cheek, banned the term” within the NCTA, he quipped.
Let’s just say it: If you’ve ever had to say “OCAP,” with feeling, in front of a human being or a camera, you’re secretly grateful. It’s just goofy-sounding.
In its place: The OpenCable Platform. (Ooo, baby.)
OK. Five months have passed since we all started culling “OCAP” from our everyday industrial conversations. It’s a mixed bag. It used to be that “OpenCable” described the hardware side of things, and OCAP denoted the software side. Now those distinctions need to be made differently. We’re hearing a lot of “what we used to call OCAP,” but that still doesn’t make it as cool as the artist formerly known as Prince...
Meanwhile, the OpenCable Platform moved front-and-center in the NCTA’s pointed response to the FCC’s Third Notice of Proposed Rulemaking this summer (see #15). And, in headends across the U.S., digital video controllers (DNCS, in SA’s terms; DAC, in Moto’s) got outfitted to do OpenCable. The boxes are moving up in the queue, too: Time Warner Cable said last month that it’s put 150,000 of them in U.S. homes already. –LE
12: EdgeQAMs: Where the action is
The first family of edgeQAMs supported digital broadcast video services, the second supported VOD, and the third group of edge products added SDV to the mix.
But a new generation of Universal edgeQAMs (U-eQAMs) emerged mid-year, and the devices support the aforementioned services, as well as DOCSIS traffic. Some examples are Vecima Networks’ HyperQAM, Cisco’s RF Gateway Series, Arris’ D5 U-eQAM, BigBand Networks’ Broadband eQAM (BEQ) and C-Cor’s CHP eQAM, which C-Cor claims is the industry’s first 1GHz eQAM.
EdgeQAMs are the linchpin of switched digital video (SDV). Every operator is either using them or thinking about it. The edge devices also enable cable operators to efficiently manage bandwidth for video and high-speed Internet applications.
Part of the rationale for the edgeQAM was to alleviate the demands on CMTS resources by assuming some of the routing responsibility (read: Make it more affordable to do video-over-IP, as well as MPEG). The concept was part and parcel of the DOCSIS 3.0 approach of the modular cable modem termination system (M-CMTS), but with the U-eQAM it appears that QAM technology is becoming more sophisticated than anyone – except QAM manufacturers – might have imagined. –TP
13: Targeted ads/’Canoe’
First of all, it’s not an acronym. “Canoe,” in mid ’07 cable conversations, was the hush-hush advertising project that everyone seemed to know about anyway. What it is: A cross-MSO initiative to make it easier for big, national advertisers to “buy cable,” even though the nation’s major TV markets are served by different cable operators. So, in one sense, “Project Canoe” is all about interconnecting the interconnects.
It’s also about interactive advertising. No groaning, please. This isn’t clicking to buy Jennifer Aniston’s sweater. It’s more like you’re watching the car ad, and it’s a car you want, so when you see the clickable thing on the screen, and you click it (using the TV remote), you go to a longer clip, on-demand.
Status: The Canoe group issued a request for information via CableLabs in August; responses were due in September. From there, a handful of finalists will submit their big plan for how to put it together.
Oh, the name? Our spies tell us it came from a comment made by Comcast COO Steve Burke, about how all cable providers needed to be paddling in the same direction. –LE
14: Comcast: Obligatory
Yeah, and USC, Ohio State, Michigan and at least one of the Florida teams automatically get slots in the Top 25 NCAA Division I football polls every year. A big-time program and consistency count.
What qualifies as big-time in cable? By the third quarter, Comcast was counting 48 million homes passed, and had accrued 24 million basic subs, nearly 13 million broadband customers and 3.7 million people taking VoIP.
During the year, it acquired Patriot Media, and simplified operations by divvying up assets shared variously with Time Warner and Insight Communications.
Comcast also bought online ticketing operation Fandango, and entered a display advertising partnership with Yahoo! From an accountant’s standpoint, what cable companies do is recurring revenue from services. Remote ticket sales and online advertising both qualify.
Meanwhile, Comcast is a big force pushing DOCSIS 3.0, the OpenCable Platform, and the transition to switched digital video.
It seems like it’s barely paying attention to an effort to develop DVRs with TiVo software, and yet that’s still enough to encourage TiVo investors. OTOH, maybe Comcast is just inoculating itself from a TiVo patent infringement lawsuit (see #37 EchoStar).
Being big makes you an easy target, and toward the end of the year, Comcast became Exhibit A in a revival of the Great Net Neutrality Debate. Comcast was accused of using deep packet inspection (DPI) technology to cut off P2P file-sharing applications, which Comcast does routinely (as do other operators) in the course of its traffic management operations. The reasoning: P2P applications will simply keep trying to
