Moto Preaches Cable's Upstream Savior
Rakib, if you don't recall, was the former chief of Terayon Communication Systems, which invented S-CDMA (Synchronous Code Division Multiple Access) as a way to boost throughput in the noisy low frequencies of cable's spectrum. CableLabs adopted S-CDMA, royalty-free, as an advanced physical scheme for Docsis 2.0.
Rakib long held that Terayon's experience with S-CDMA would give it a big advantage over the competition, but those ambitions were stymied when the vast majority of cable operators opted to use A-TDMA (advanced time division multiple access) -- the other advanced physical layer defined by Docsis 2.0.
Now Motorola, which acquired Terayon in July 2007, is knocking the dust off S-CDMA, touting it as a technique MSOs should employ to clean up and enhance their upstream spectrum before they try to offer faster speeds via Docsis 3.0 channel bonding. (See Motorola Seals Up Terayon and Motorola to Buy Terayon for $140M.)
Finishing 500 laps
S-CDMA can rapidly address the need to maintain higher average upstream throughputs, letting operators apply channel bonding later to get higher peak speeds, claims Floyd Wagoner, the director of marketing and communications of Moto's access networks unit.
"Channel bonding is a race to the finish line without completing the 500 laps," Wagoner says. [Ed. note: Or 200, if you're talking about the Indianapolis 500.]
Moto claims that deployment of S-CDMA, at "low single dollars" per node, betters the economics of node splits and a new breed of CMTS cards that increase upstream port densities. Javelin Innovations (née Vyyo Inc.) is working on new upstream-centric design that taps spectrum above 1 GHz, but it's been on radio silence and hasn't revealed any recent pricing info. (See Moto Wields Upstream CMTS Blade, 'New' Vyyo to Rise in March , and Vyyo's New Name & Game .)
Although channel bonding will eventually enable MSOs to offer shared upstream speeds greater than 100 Mbit/s, Motorola thinks the use of S-CDMA in single-channel Docsis 2.0 environments will give MSOs the speed they need today to stay competitive. (See Verizon Assails Cable With Amped Upstream and Qwest Attacks Comcast With 40 Mbit/s.)
(By the way, Moto isn't doing upstream channel bonding in the field yet with the BSR 64000, its flagship CMTS, but it's been doing downstream channel bonding with multiple MSOs for some time.)
Moto claims S-CDMA increases upstream capacity by as much as 50 percent, but how does that translate into speed? The company has been able to generate a consistent 25-Mbit/s flow through one 6.4-MHz channel using 64 QAM modulation, according to Mike Cookish, director of product management for Moto's access networks division. And that's using a channel in the noisy spectrum below 20 MHz that's typically unusable with A-TDMA, he says. Any channels recovered in that "barren land" using S-CDMA can be applied toward Docsis 3.0 upstream channel bonding.
Motorola is going around showing MSOs how they can configure their CMTSs and modems to support S-CDMA, but it's got no deployments to brag about yet. Cookish notes that some MSOs in Europe with older plants with high levels of ingress noise in the upstream are among those interested in conducting S-CDMA trials.
An engineering exec with a top-five U.S. MSO agrees that bandwidth is scarce in the lower frequencies -- 5 MHz to 16 MHz -- and that S-CDMA "performs better in that band."
So why haven't MSOs adopted it? It wasn’t because they weren't eager to use S-CDMA. There's just been a historical lack of vendor support, except from Terayon (and now Motorola), at least from the standpoint of having access to equipment with S-CDMA on board that's been tested and approved at CableLabs.
"The interest was there... The product wasn't available until recently," the cable exec notes.
Another industry observer familiar with the history of the technology says chips with S-CDMA activated remain unproven in the field and the "impact of S-CDMA on cable operations is still unknown."
Checking the competition
Motorola's perceived advantage with S-CDMA may be short lived, considering what's already underway among its CMTS rivals.
Arris Group Inc. (Nasdaq: ARRS) has supported S-CDMA in its Docsis 2.0 releases since 2005, and expects to do the same via an "imminent release" for its Docsis 3.0 gear, a spokesman says.
Cisco Systems Inc. (Nasdaq: CSCO), whose primary CMTS product is the ubr10012, will offer the S-CDMA feature "in our next software release," a spokeswoman says, without indicating the exact timing.
Casa Systems Inc. , an Andover, Mass.-based startup, claims to have the only CMTS with S-CDMA upstream channel-bonding capabilities today. Casa, according to CEO Jerry Guo, has that ready to go in three models -- the c2200, c3200, and chassis-sized c10200. (See CableLabs Cheers Casa Chassis.)
Activating S-CDMA in those products is strictly up to the MSOs, but thus far all of Casa's customers have chosen to deploy its gear with A-TDMA. "As far as I know, no major MSOs in the world are using S-CDMA," Guo says.
Although Moto's recent S-CDMA evangelizing appears to downplay cable's near-term need for upstream channel bonding, the company says the latter remains firmly on its CMTS roadmap. "Cable needs to have 100-meg in their sights," Cookish says of cable's future upstream capabilities. But obtaining the necessary channels to bond in cable's thin upstream will be made easier if S-CDMA is used to clean up and optimize the noisiest part of cable's spectrum, he maintains.
At least one MSO is already trying out upstream channel bonding: Japan Cablenet Ltd., the second-largest MSO in Japan, in conjunction with Arris. (See Japan Cablenet Swims Upstream .)
— Jeff Baumgartner, Site Editor, Cable Digital News