Same rules as the last part apply, generalizations will be made, engineerï¿½s will have their toes stepped on, you are welcomed to nit pick in the forums.
It has to be good to be Qualcomm today, sitting on the top of the CDMA world watching everyone knock on their door, converting to CDMA, and having carriers ask left and right to change their plans, and not the other way around. But Qualcomm has finally put all the pieces together to assert dominance in the next decade to come, something to set Qualcomm CDMA apart from a world dominated by CDMA, and differentiating from emerging technologies (such as WiMAX and Flarion).
When Qualcomm was creating EV-DO, they saw the GSM world developing W-CDMA. CDMA2000 uses 1.25 MHz of spectrum per carrier channel, W-CDMA (or UMTS) uses 5 MHz. Thatï¿½s a huge difference today. The initial impulse would have been to simply start stacking 1xRTT streams on top of each other, similar (but much more functionally sound) to putting two dial-up modems together. While that technology didnï¿½t work well, Qualcomm has long established that a 2xRTT would have, however, at the significant cost of spectrum. 4xRTT (no, that wasnï¿½t ever made, but could have been as an alternative to EV-DO) would have run at around 576k, far above UMTS and using the exact same spectrum as UTMS. But, Qualcomm saw a better way.
While Code Division works great for voice, we canï¿½t exactly rebuild a chunk of data that gets lost in a tree or brick wall. You see, CDMA works by rapid firing tons of small packets of information over the available spectrum, and then rebuilding that stream at the tower (thatï¿½s why time synchronization is so key to CDMA). In voice, if a part of the call gets lost, itï¿½s simply rebuilt using parity technology in the decoder at the tower. This also explains why people complain that CDMA calls sound ï¿½roboticï¿½, little do they know, that if the same environment had blocked a TDMA/GSM stream, nothing at all would have been heard. So, Qualcomm sought out to take CDMA and stream it through a time divided data stream. So while EV-DO uses TDMA-style air frequencies, it is still using the principles found in 1x and CDMA. The positive effect is that spectrum usage is still as low as in traditional 1xRTT, 1.25 MHz stays 1.25 MHz. The problem is that towers have to segment off spectrum in terms of hard boundaries, EV-DO spectrum allocations cannot be shared by CDMA voice calls. In 1xRTT, everything is shared (though with QoS, towers can protect from data impacting voice calls at the carrierï¿½s discretion).
This brings us to EV-DV. EV-DV essentially brings the TDMA data stream of EV-DO back into code division. Through advances in the CDMA platform, EV-DV can now transmit at even higher speeds than TDMA. To put it in simple terms, Qualcomm figured out how to make packets of data getting lost in a tree less of an issue, while at the same time improving bandwidth speeds even in optimal environments.
We waited for to publish this article to put a bit of a bombshell in context; Sprint will likely officially drop support for EV-DV. But why? After all, Sprint held off for EV-DO in order to get equipment that was software-upgradeable (or ï¿½free after rebateï¿½) to upgrade to EV-DV, and signed a major deal with Qualcomm to ensure this. This is because EV-DV will change, not the underlying goals, but how it will be marketed and what can be upgraded to support it.
To put the new path for EV-DV in a sentence is brutally easy: merge everyone under one platform, sacrificing total dominance over the platform, but making it the global, and universal standard, while at the same time as offering competition, allowing their own chipsets to remain the best in the business. This is a three phase strategy: make Qualcommï¿½s CDMA not seem superior, make the competitions CDMA interoperable with Qualcommï¿½s, then deliver a unified forward migration path from UMTS and EV-DO to one platform.
The advantage to merging platforms is clear more in America than anywhere else. Today, you have to have a phone for TDMA, a phone for CDMA (which can use AMPS, as can TDMA), a phone for EDGE, a phone for iDEN, and a phone for GSM and UMTS. That comes to five cell phones, three assuming the rollup of TDMA, AMPS, and iDEN. Qualcomm is the one company in the industry that is willing to sacrifice dominance over its own platform in order to deliver a world where one phone can use every tower regardless of location or platform.
Probably one of the biggest hurdles to EV-DVï¿½s assertion over the GSM world is the lack of an easy migration path from UMTS to EV-DV. EV-DO has a clear migration path, some of Sprintï¿½s towers can theoretically upgraded to EV-DV with a firmware update (however, with the expected changes to EV-DV underway, this is unlikely to be possible). Also a problem is EV-DVï¿½s lack of support for GSM vocoders. If EV-DV supported existing GSM-world technologies as best as possible, it is quite possible for Qualcomm to market this technology.
Itï¿½s also important to realize that Qualcomm probably will not call EV-DV, EV-DV. Once these modifications to the platform (incorporating GSM vocoders and creating some sort of a forward migration path from UMTS/HSDPA), Qualcomm will abandon all references to EV-DV. But, the core goal of the new platform, a CDMA powered by an EV datastream reincorporated into CDMA airwaves will remain the key functional advantage.
Qualcomm will also have a backup plan. With only 1.25 MHz, if HSDPA gets upgraded, all Qualcomm has to do is use a second, third, or even fourth data stream. Again, UMTS is 5 MHz, Qualcommï¿½s bar is significantly lower.
So, the question is begged: whatï¿½s the proof? First, is Qualcommï¿½s recent affinity to the GSM worldï¿½s CDMA platform, UMTS. It makes no sense otherwise for Qualcomm to back technology that they have previously labeled as inferior, reverse engineered, and less secure. However, if nobody knows CDMA better than Qualcomm, it is easy to pick Qualcomm as your chipset manufacturer. Qualcomm delivers UMTS better than the people that developed UMTS, and Qualcomm looks like the good guy. Second is Qualcommï¿½s want to unify existing technologies under a single chipset. The MSM6700 provides support for EV-DV, EV-DO, GSM, CDMA, and CDMA2000. It supports every U.S. platform except for Analog, TDMA, and iDEN, all of which are currently begin phased out. Having all carriers sign roaming agreements, even across platforms, is necessary for a unified platform to seem clear to all sides.
The benefits are endless to everyone. Instead of regulations deciding platforms, capitalism will. The tidal wave of the merger of the GSM world and the CDMA world around a single platform would dominate cellular communication for as long, if not twice as long, as analog did. Consumers wouldnï¿½t be last-place either. With each side offering some tradeoffs, you will see phones that will work on any carrier, SIM cards on CDMA, common platform, with networks switching between Qualcomm-patented vocoders and GSM-controlled vocoders completely seamlessly. GSM die-hards will even be offered the option of simply not including Qualcomm vocoders, while this would prevent some network interoperability, phone manufacturers will most likely limit such offerings to low-cost phones that would not be impacted by total network interoperability.
The timeframe, however, for this is debatable. But it is clear that the GSM world is preparing (beginning with W-CDMA) to support CDMA, either in an attaching to a unified global standard for cell phones, or a competing CDMA that uses GSM codecs with CDMA functionality. This is all due to Qualcommï¿½s 17-year long patents on the un-clone-able portions of CDMA will begin to end. It is quite possible that even before the MSM6700ï¿½s announcement that GSM developers realized this possibility and have begun work on different means to a similar end.
Hence, the biggest question at this point boils down to is the GSM world working on a CDMA to cell to Qualcomm, is Qualcomm working on a form of EV-DV to sell to their new best friends in GSM, or are both happening at the same time? The chances of any of the three are much more probable than two competing, comparable and similar CDMA technologies existing at the same time.