How Cable Gets More Stuff Upstream

So far, we've looked at the types of traffic that flex cable's upstream path (March 19), and how its slimness and location make it sensitive to discord (April 2).

Now: What to do about it?

There are two main ways to trick the upstream into carrying more stuff. One is to split the node, which halves or quarters the number of homes that share an upstream ride.

Most current optoelectric node equipment comes with four output legs. Each leg connects to the coaxial portion of the network (destination: homes). Four ports means two potential splits, say, from 500 passings to 250 passings, then from 250 passings to 125 passings. (Note that it's never that exact.)

Node splitting, MSOs say, takes about a half-day and costs around $10,000, including labor and equipment (mostly lasers and transmitters/receivers).

The second means is to use what engineers call "higher-order modulation" in devices that rely on the upstream path, like cable modems, digital set-tops (for interactive sessions) and phone gear.

Higher-order modulation is tech-speak for "better than what you use now."

As previously discussed, today's gear uses modulation called "QPSK," for quadrature phase shift key. It's is a bruiser — stalwart, but steamroller slow.

Technically, QPSK can still usher data when carrier-to-noise ratios dip down to 9 decibels (dB) or so. But traffic moves sluggishly (relative to better modulation), at about 1.2 Mbps per MHz.

Carrier-to-noise is a commonly used measure of plant performance, expressed as a ratio of the power of the carrier signal, to anything that impedes that carrier signal. It's stated in decibels. Lower is bad or cruddy. Higher is good or clean.

The at-bat form of upstream modulation, already aboard most cable-modem chips that are currently shipping, is 16-QAM (Quadrature Amplitude Modulation).

It's faster, but requires a cleaner path: 16-QAM can only shimmy down to around the 16-dB range, in carrier-to-noise ratio.

But it moves bits faster — up to 2.5 Mbps per MHz, or about double what QPSK can do. (Note: While 16-QAM is on-chip, it hasn't been widely activated anywhere.)

In the "good" upstream spectral zone, between 20 and 45 MHz, 16-QAM provides about 62.5 Mbps of aggregate throughput, to QPSK's 30 Mbps.

Coupling 16-QAM in the upstream with a single split of a 500-home node gives some subset of 250 homes passed shared access to 62.5 Mbps. Half the homes share twice the speed.

The size of the subset depends on service penetration, the number of devices simultaneously sending data upstream and what those devices are doing.

That's much better than some subset of 500 homes-passed sharing 30 Mbps of aggregate upstream throughput.

And there are even higher forms of higher-order modulation on the near-term horizon. The Cable Television Laboratories Inc. DOCSIS (Data Over Cable Service Interface Specification) group has long been evaluating modulation techniques that make the upstream ride either faster or sturdier. Preferably both.

Once upon a time, this work was unofficially called "DOCSIS 1.2." Then, it was officially changed to "Advanced PHY," for Advanced Physical Layer. "Physical layer" means anything to do with moving information over the cable plant.

In small engineering circles, the blare about advanced PHY is nearly as noisy the upstream path itself, except that it's fairly predictable, because the hullabaloo is happening between vendors. Specifically, between chip-set vendors.

Indeed, next-generation upstream modulation, in the context of technology politics, brings to mind a recent cartoon in The New Yorker. Two men are talking. Caption: "When you say campaign reform, is that with or without an asterisk?" Applied to the upstream, the image could've just as easily read, "When you say advanced PHY, is that with or without an asterisk?"

The asterisk is the whole messy history about: No. 1, which upstream technique is best; No. 2, which is backwards-compatible with all the stuff that's already been deployed; and No. 3, which chip vendors made which blustery claims about their technology's role in the ultimate advanced-PHY specification.

That decision, notably, hasn't yet been made by CableLabs' MSO-populated Certification Board, but is tentatively scheduled to happen in the next six months.

It's tentative because chipmakers, predictably, are tendering massive technical arguments and counterarguments. At the root of it is Terayon Corp. and its "S-CDMA" (Synchronous-Code Division Multiple Access) modulation, versus a competing technique called "FA-TDMA" (Frequency Agile-Time Division Multiple Access) proposed by Broadcom Corp., Texas Instruments Inc. and others.

The technical differences between S-CDMA and FA-TDMA are way too complicated to succinctly translate in the second-to-last paragraph of a column. Boiled way down, both techniques claim to dramatically increase upstream throughput, even in the face of big noise.

Terayon says its method is backwards compatible to DOCSIS 1.0 and 1.1, approaches 128-QAM speeds and does better with noise. Broadcom touts 64-QAM speeds and backwards compatibility. Both claim to upshift and downshift around noise.

Whatever the ultimate decision, advanced PHY is probably a 2002 thing. 16-QAM is a soon thing. Node-splitting is a now thing.

All together, they go far in assuaging cable's upstream dearth.