CDMA is dead: Long live OFDM. That is the cry of a growing legion of OFDM supporters who feel the next generation of mobile networks has no choice but to gain spectrum by using orthogonal frequency-division multiplexing technology. But here's
CDMA is dead: Long live OFDM. That is the cry of a growing legion of OFDM supporters who feel the next generation of mobile networks has no choice but to gain spectrum by using orthogonal frequency-division multiplexing technology. But here's the rub: Will OFDM be used to enhance existing networks or will it, like WCDMA, require the build-out of all new networks? Furthermore, the timing carries a huge question mark.
Groups working on what they call Super 3G and 4G lend support to the belief that the writing is on the wall for CDMA. Recent changes in 3G development may have added more credence to that theory. Qualcomm has ceased development work on 1xEV-DV technology, and Texas Instruments appears to be doing the same. US operator Sprint, which was steadfastly dedicated to waiting for 1xEV-DV before grudgingly announcing it would deploy 1xEV-DO later this year, has officially abandoned its 1xEV-DV plans and will instead upgrade to 1xEV-DO Revision A in 2006-2007.
Perry LaForge, executive director of the CDMA Development Group, said 1xEV-DV's turn of fortune reflects carrier preferences in light of the fact that 1xEV-DO has progressed quite rapidly in terms of capabilities and deployment. "Almost all of the [CDMA2000] operators are considering DO, moving to DO or have already implemented it," he said.
Yet others see a deeper message in the retreat from 1xEV-DV. It could be that operators feel 1xEV-DO, HSDPA and HSUPA represent the upper limits of today's CDMA technologies. In turn, the door may be opening to integration OFDM - which splits a radio signal into multiple narrowband channels at different frequencies - and OFDM with MIMO - a complex approach that requires multiple antennas on the transmitter as well as the receiver. Or perhaps OFDM with MIMO will form the basis of brand new non-CDMA networks, as some proponents of WiMAX envision.
Industry analyst Peter Rysavy said operators must correctly time the crossover point from CDMA to OFDM. "OFDM-based solutions are not mature whereas the CDMA-based systems are. So, that's the dilemma. When do you make that transition? If you do it too early, you're going to experience a lot of pain," Rysavy said.
There are nascent efforts to integrate OFDM into existing WCDMA radio networks via an envisioned standard for High Speed OFDM Packet Access. HSOPA would potentially offer download capability of 40 Mbps.
The 3GPP also has MIMO on the roadmap for WCDMA as part of a Release 7 feature. "I don't think it's clear to the industry that taking this step is going to be the right step," said Rysavy, explaining that applying MIMO to CDMA rather than to OFDM is computationally much more difficult. 3GPP initiated a study group in late 2004 to investigate this.
"The cellular industry right now is getting a little nervous because they're not sure that the current roadmap is going to have them be competitive three to five years from now," he added.
One looming threat comes from WiMAX, an OFDM technology that will use space-time coding (STC) has drawn considerable attention, with some predicting new operators using this technology will unseat today's mobile phone networks. "Three years ago we went though the same discussion regarding whether Wi-Fi was going to replace cellular networks," LaForge observed. "Things tend to get oversold and over-hyped in terms of their capabilities. We saw that with Wi-Fi, we're starting to see it with WiMAX. WiMAX will settle down at some point into a role for itself. What that role is, I don't think people really know yet."
According to LaForge, "You'll see CDMA continue to develop, but it'll be augmented longer term with various approaches, which might be OFDM-type technologies." He also noted that there will be opportunities for operators to use different technologies at several frequencies, with devices accessing them for as needed.
OFDM vendors such as Flarion Technologies and Alvarion have made names for themselves in this arena, but actually all major network vendors are getting involved with OFDM-based systems.
John Hoadley, vice president of advanced technology-wireless networks at Nortel Networks, which has been aggressive in developing OFDM technology, expects the migration to OFDM/MIMO will occur in steps. For example, the WiMAX mobile standard - 802.16e - will likely be ratified this summer but will initially be applied only to nomadic computing rather than highly mobile data services. "A full OFDM/MIMO system that handles users at high speeds, that's probably a few years off. But I think we'll see a gradual progression toward that as we go along," he said.
Hoadley predicts that 3G can expect a relatively long lifespan thanks to advances expected from 1xEV-DO Rev. A and HSUPA, which will drive adoption of broadband. But they could become victims of their own success. He notes, "I see the critical event driving OFDM/MIMO being those [3G] successes happening first and then operators trying to make things more efficient."