LTE is not the new FTTH - yet

One of the more notable moments at LTE Asia happened during the opening keynotes when Eric Gan, president and chief operating officer at Japanese cellco eMobile, said that he viewed LTE not only as a data play, but also as a viable substitute for FTTH.

In fact, Gan described mobile broadband "as a continuation of ADSL, not FTTH", and that LTE would allow eMobile to compete even against Japan's growing FTTH market.

"We have performed tests with LTE, and the speeds are getting closer to fiber speeds," he said. "That's why we believe LTE will cannibalize the fiber market."

That touched off some discussions on and offstage as to whether eMobile was putting a little more faith in LTE's capabilities than even the most optimistic expectations of the technology warrant.

While most people I talked to at the show agreed that LTE could certainly make ADSL obsolete and even serve as a viable replacement (at least in service areas where VDSL2 isn't a realistic option, or where copper doesn't exist at all), they were much more skeptical of the idea that LTE could eat fiber's lunch, if only because fiber in Japan was simply too widely deployed and too fast for LTE to realistically compete with it.

And while one vendor did put up a slide claiming that HSPA/LTE will account for 80% of the global broadband market by 2014, an argument could be made that a huge percentage of those will either be ex-DSL users and/or people where fiber (or indeed broadband of any flavor) isn't yet available and won't be for another ten years, if ever.

So it seems fair to say that in terms of throughput speeds, LTE won't fare well as a fixed-broadband substitute where FTTH is already in place - especially if FTTH providers are using it to deliver very heavy video services like, say, 3D high-def television.

LTE Advanced, on the other hand, could be a very different story.

LTE Advanced - which is being developed by the 3GPP to conform with the ITU's IMT Advanced specs - essentially promises to take LTE's throughput speeds up to 1 Gbps (downstream) and 500 Mbps (upstream), as well as double throughput performance at the cell edge and lower latency on the control plane. Just how it will do that is still being worked out, though a number of ideas are on the table, such as widening the channel path from 20 MHz to 100 MHz and use of relay nodes, to name just two.

Granted, all of this is years away. The ITU has just started sifting through half a dozen proposed technologies for IMT Advanced, all of which are centered around either 3GPP Release 10 (LTE Advanced) and 802.16m (Wimax), and is slated to make its final decision in October 2010. Commercial deployments after that will depend greatly on both LTE's progress and - more importantly - regulators finding several hundred MHz worth of spectrum in a harmonized band to set aside for it.

Assuming the ITU will create another IMT family of standards rather than picking just one, that also creates yet another opportunity for Wimax as well. For all the hand-wringing over Wimax' low market share compared to LTE, Wimax already serves as a DSL substitute where it's been deployed. Via Wimax 2.0, it may actually have a second chance to take advantage of the upgrade lull as cellcos take their time upgrading to LTE anywhere between now and 2020 - although Wimax 2.0 will be subject to the same spectrum availability issues.

In any case, once such issues are resolved, and even assuming that ITU Advanced's real-world throughput will be at least half its peak speed (let's be really conservative and say 300 Mbps), mobile broadband could indeed potentially give FTTH some direct competition, not least by being mobile.