How much life is left in QAM technology?
Some 20 months ago, at Paul Kagan’s “QAM Before the Storm” event that preceded the SCTE Conference on Emerging Technologies (ET) in Los Angeles, one proponent of universal edge QAM devices noted “wistfully” (to borrow the bon mot used by Heavy Reading Senior Analyst Alan Breznick) that given the fundamental economics of the two technologies, RF is in a losing battle against photonics.
In other words, the shift from RF over coax to IP over fiber is inevitable. And if that is the case, then RF over glass (RFoG) certainly looks like an intermediate step.
Two technical papers
At this year’s Cable Connection week in Washington, D.C., however, two technical papers revived RF’s standing as a contender. As it happens, they both had the same author, Dr. Robert Howald, customer systems architect in Motorola’s office of the CTO.
In “Fueling the Last Coaxial Mile,” a paper presented during the ET portion of the April event, Howald explored unused capacity of the HFC plant. As an “interesting reference point,” he made this observation: “Note that 1 GHz of ideal 1024-QAM bandwidth, at 10 bits/s/Hz efficiency, adds up mathematically to 10 Gbps.”
That’s a lot of throughput, comparable to what’s promised with emerging 10Gig passive optical networking (PON) recommendations from the IEEE and ITU-T. But 1024-QAM? Didn’t Howald explain the challenges of deploying 1024-QAM in a Cable-Tec Expo paper back in 2002?
Yes, he did. Which is why it’s important to digest “New Megabits, Same Megahertz: Plant Evolution Dividends,” the contribution from Howald (and colleagues Amarildo Vieira and Michael Aviles) to the NCTA Technical Papers this year.
“The good news,” conclude the authors, “is that interference analysis conducted in 2002 based on findings extrapolated from 256-QAM performance of that era now seems to have been addressed in the receivers available today.” They continue:
“Given the results shown here, an awareness of the key drivers to link performance, and a modeling approach that can be used to assess HFC readiness to accept 1024-QAM signals, operators can start netting themselves that extra 25 percent of bandwidth efficiency that is currently going unused in already occupied spectrum.”
To some, this may sound like QAMunism, i.e. the theory and practice of constantly moving to ever-higher orders of modulation. One reality check, for instance, are ongoing challenges that operators have with 64-QAM in the upstream. (For more, stay tuned for next month’s webcast on the topic.)
Yet it’s pretty hard to dismiss the author of that memorable paper from 2002, who returned with two more brilliant ones this year.