Migrating toward an all-digital world
SCTE’s ’04 Conference on Emerging Technologies explored the
all-digital options and dissected the potential potholes ahead
The Society of Cable Telecommunications Engineers certainly picked an appropriate site for the 2004 Conference on Emerging Technologies. "Big D" usually stands for 100 percent Dallas, but, as conference-goers soon discovered, it also stands for 100 percent "Digital," too.
Going all-digital was the centerpiece of this year's ET, with presenters and panelists describing and debating how to get there, the costs involved and the challenges to be faced.
Although ET's goal is to get a solid glimpse of the technological developments that could impact the industry three to five years into the future, the window for cable's all-digital migration might be even tighter than that.
Just one day after the conference, Charter Communications emerged as the first MSO to reveal details about its migration plans, announcing it had flipped the all-digital switch in Long Beach, Calif., a system that supports about 188,000 homes passed.
There, the MSO is delivering analog-free programming via two primary elements: a digital compression system that encodes the entire program line-up and a digital program insertion system. Harmonic Inc.'s DiviCom MV 50 variable bit-rate encoders and DiviTrackXE closed loop statistical multiplexing system are powering the all-digital system.
Charter is tackling the migration with a simulcast system designed not to disrupt existing analog subscribers. "If you have a digital box, you will have 100 percent digital programming. If you don't have a digital box, nothing changes," explained Wayne Davis, Charter's senior vice president of engineering and advanced technology. The idea is to increase digital box penetration over time until all customers are on the all-digital platform.
Though the benefits of all-digital extend to reclaimed analog spectrum for high-definition television and video-on-demand, it also improves picture quality significantly. "When we receive off-airs from antennas, unless it's a direct feed from the broadcast, that doesn't guarantee you're getting a good quality picture," Davis said, noting that 20 of the 96 analog channels offered in the market are off-air. By going all-digital, carrier-to-noise is reduced and distortion is eliminated, he added.
To further its all-digital effort in Long Beach, Charter is also looking to deploy special dongles and stripped-down all-digital set-tops, but wasn't ready to disclose specific details about those plans. Long Beach's digital network is presently based on the Scientific-Atlanta platform.
Back at the conference, Tom Staniec, vice president of network operations and engineering at Time Warner Cable, kicked things off by mapping out some of the critical technology issues for the cable industry in an all-digital future.
He told the crowd the move toward all-digital may be as much evolution as it is revolution.
On the revolution side, Staniec noted that how cable operators think about simple things like channels may have to change.
"One question you have to contemplate is, do I need channelized 6 MHz channels anymore? Or do I need to deliver one Gig to each house?," he asked.
Capacity, too, would be an issue in the all-digital future. Staniec noted most cable operators that built out to 750 MHz are now finding that might not be enough.
But he also argued that cablers don't have to necessarily drill that high capacity out to every potential customer, nor does he think fiber to every home is necessarily the answer.
Instead, it may lie in future products such as a "lambda on a pole"–a single-point fiber extension that could ride next to the existing network and service just one customer. While the network technology may be developing to offer these future higher-speed capabilities, a critical roadblock will be developing the back office systems to monitor network traffic and bill customers for the service.
"You have to start thinking about this because you start putting together these systems for the future, [and] you may be billing in a way you never have before," he said.
customer. Black lines are the cost per customer at various penetration levels
assuming traditional 750 MHz upgrade. Gray lines are cost/customer for a
digital upgrade assuming a $70 digital set-top, 2.5 set-tops per home
and a $30 set-top install fee/customer. —Source: SCTE, Cequel III.
"Going all-digital is a competitive response," said Don Loheide, vice president of network engineering for Cequel III, noting that DBS providers have become very good at explaining that they, not cable, are the ones that currently offer 100 percent digital programming.
And that's not all. RBOCs such as Verizon and BellSouth will apply even more competitive pressure when they build and activate their next-gen fiber-to-the-premises networks.
Going all-digital also offers its share of operational efficiencies, because it enables the operators to eliminate headends and potentially reduces service theft and truck rolls, Loheide said. And, by adding a cable modem to the picture, operators will also gain an intelligent device that can provide plant condition information and help with targeted maintenance.
The benefits also extend to cost savings (see chart, above). Under one model Loheide presented, a cable operator with 60 percent digital penetration could save between $195 and $325 per customer.
But going all-digital isn't without its challenges. Operators, he explained, can't just migrate channels to the digital tier if they are contractually bound to analog. Additionally, some networks aren't available in digital format yet and would have to be encoded locally.
Regardless of the challenges it faces, the cable industry must establish digital as the new standard for cable, said Greg Thompson, chief architect of Cisco Systems' video networking business unit.
But, he warned, cable must avoid being locked into MPEG-2. Though it would be nice if all-digital boxes could support new codecs in software, it will come at a penalty of additional memory, Thompson pointed out.Transporting digital
An afternoon presentation round featured a raft of vendor representatives–each presenting their own best scheme to handle core transport on these future all-digital, multiservice systems.
Based on the presentations, that "best network" depends largely on whom you talk to.
A. David Guiterrez, senior product manager at Fujitsu Network Communications Inc., made the argument for Sonet fiber networks as the core of this future all-digital network. Already an established metro transport technology, he argued it provides a stable platform–and boosted by add-on technologies such as Multi-Service Provisioning Platform–it is a strong choice for the job.
But rings weren't necessarily the best network layout for David Lively, senior manager of marketing at Cisco Systems Inc. He argued that cablers should consider re-establishing the older coax spoke and hub design in fiber, particularly to handle the downstream heavy/upstream light video-on-demand services now starting to crowd the broadband pipes.
That doesn't fit well with unidirectional ring networks, Lively pointed out. Hub and spoke, meanwhile, can be made 100 percent efficient in theory, with content sent only one direction down a branch.
Meanwhile, optical Gigabit Ethernet was the way to go for Graham Smith, cofounder and vice president of systems engineering for Internet Photonics Inc. He argued optical Ethernet–in effect running Ethernet over DWDM lambdas–could perform much like a tried-and-true Sonet fiber network but without the cost and operational problems.
GigE without the extensions for optical transport "is not carrier grade to handle the transport," he said.
or the way to the all-digital goal. Source: SCTE, Cisco and nCUBE.
Java: iTV's new onramp?
With a slew of operating systems and conditional access schemes already deployed, finding a way to create a single platform able to deliver applications uniformly is no small challenge. So Steve Calzone and Eric Miller pitched an idea to unify digital using the Java programming language as an "onramp" for development of common interactive applications in an OpenCable Applications Platform-run future.
Calzone, iTV systems engineer at Cox Communications, noted while more advanced digital boxes being deployed today come close to OCAP compatibility, there still is a large installed base of older boxes, such as the Motorola DCT2000, that can't conform to OCAP. Developers, meanwhile, face the frustrating problem of having to design multiple versions of their software applications to fit the many box flavors.
"We need to resolve this problem. We need to come up with a common development platform." Calzone noted.
Miller, CTO at Vidiom Systems Corp., picked up on that theme, saying "while we are encouraged by all of the current work in standards, we realize there is a missing piece, and that is what we will propose here."
What the duo proposed was an extension to OCAP that makes Java, the Internet programming language developed by Sun Microsystems, the uniform platform for applications development. Designed especially for the Internet's distributed environment with simpler coding than other languages, Java already is part of the OCAP specification.
Calzone and Miller proposed creating a Java-based applications protocol interface set simple enough to access boxes such as the DCT-2000 while supplying some of the sought-after applications found on higher-level boxes. Having this single API would greatly simplify applications development and interactive services deployments for MSOs.
Calzone said he and Miller have been talking with other MSO and gear vendor representatives about the idea, adding he was hoping for input from the conference "to really move this thing and make it happen."
Miller added that with satellite providers rapidly developing the next round of digital applications and services, "we can't wait for it for five years, so there does have to be a little bit of a fast track."
Another idea gaining some attention these days involves harnessing the Data Over Cable Service Interface Specification (DOCSIS) channel to carry digital set-top box out-of-band signaling functions. It is the basis of the DOCSIS Set-top Gateway (DSG) specification soon to be released by CableLabs. A follow-up advanced mode specification also is in the works.
Digital cable systems use out-of-band signaling to relay conditional access and encryption commands between the headend and set-top boxes. At present, proprietary signaling systems are used to carry these messages.
Unlike the cable-modem centric DOCSIS specification, DSG's scheme would be one-way, as are the proprietary out-of-band signaling systems in existing set-tops, said John Chapman, Cisco Systems Inc. distinguished engineer. Chapman said that was necessary to make it compatible with the real-world digital systems already in place.
"If you want to take your proprietary signaling messages from their existing channel, you take the messages and up them on top of DSG," he noted. "It is completely transparent–it is a layered transport."Attendance jumps, L.A. on deck
Providing further proof that 2004 could be a turnaround year for the cable tech sector, attendance at ET '04 was solid, jumping 15 percent to almost 800, versus 2003's totals. Still, the conference has some ground to gain if it's going to catch up to recent higher water marks during the dot-com boom. ET 2001, for example, drew about 1,100.
ET will get a chance to equal or better that mark in 2004, when things get underway again in Los Angeles. SCTE has tapped cable vet Tom Jokerst to chair the program subcommittee for next year's gathering. Jokerst is presently the chief technology officer of video-on-demand server startup Broadbus Technologies. Previously, he was senior vice president of advanced technology and a senior technical advisor for Charter Communications.
"I'm looking forward to the challenge of building a program for discussing innovations and advanced technologies that are three to five years out," Jokerst said in a prepared statement.