Verve,
Re: QCOM - 3GPP2 cdma2000 v. 3GPP UMTS Decision Process
I said:
"Irwin might think that cdma2000 is the "better one", but the carriers aren't buying it, and there are reasons for this."
I do not claim to know whether W-CDMA is "better" than cdma2000 or whether cdma2000 is "better" than W-CDMA. I leave that to engineer or Clark, or Walt, or w, or Quincy, or mightylakers to debate, and it is unfortunate that we have few posters to the SI boards that can represent the W-CDMA point of view.
You asked:
"Why do you think WCDMA is giving CDMA2000 such a back-alley beating?"
"What do you think the carriers would tell us if we asked them this question?"
I'm not sure what the carriers would tell us, but these are my opinions about some of their possible decision points. One , some, or many of these (depending on carrier) may have tilted the balance sharply in favor of UMTS for 3G technology selection:
1.) Qualcomm has always maintained that given the efficiency of cdma2000 that 3G requirements can be met and benefits realized in existing spectrum using a 1,25 MHz channel, synchronously. cdma2000 as currently standardized is optimized to deliver 3G benefits in current spectrum.
2.) Proponents of W-CDMA maintain that the best way to optimize CDMA for IMT-2000 spectrum set aside in the 2 GHz range is to use a channel size of 5 MHz or greater, and that asynchronous operation which requires no special synchronization between cells, supposedly makes placing base stations in underground subways, in tunnels, and inside buildings much simpler.
3.) Theoretically both cdma2000 and W-CDMA offer twice the capacity of IS-95 A/B, and 7 times that of GSM (voice). On paper, this is a push even if in implementation it may not be.
4.) The reality is that many countries have set aside large chunks of contiguous spectrum in the 2G HZ range. Spectrum gets licensed. If you are a carrier you stand up and go for it, or you lose an opportunity to obtain a priceless and precious resource forever. If you go for it, you are generally required to make a technology decision (before or immediately after) and build it out as rapidly and as intelligently as possible. Those carriers that perceive that W-CDMA is "really" better optimized for 2G Hz IMT-2000 spectrum than cdma2000 are likely to choose W-CDMA.
5.) The 3GPP 'Release 99' (initial phase) IMT-2000 standard approved by the ITU in April was (to the best of my knowledge) compliant with the 3 basic data transmission requirements of IMT-2000 (up to 144 kbps in full mobility, up to 384 kbps while moving and up to 2Mbps throughput for fixed applications). Final stages of commercialization commenced immediately by the majority of the worlds vendors immediately upon publication.
6.) the 3GPP 'IS-2000A' (initial phase) IMT-2000 standard approved by the ITU in April was (to the best of my knowledge) not fully compliant with the 3 basic data transmission requirements of IMT-2000. Instead it offers up to 144 kbps in full mobility. Up to 384 kbps while moving and up to 2Mbps throughput for fixed applications are being addressed in later phases and in fact 1xEV-DO (which in conjunction with 1xMC meets the IMT-2000 requirements) recently became a TIA standard. 1xEV-DV awaits standardization and both await ITU approval.
7.) There is some question about whether or not the world needs yet another "interim standard" or data overlay like GPRS, or EDGE. or 1xEV-DO. 1xEV-DV may be the "real deal" once standardized next year. 1xRTT/1xEV-DO may not be seen as the "real deal", only an expedient. Sometimes sooner is "better" than later but "better". 1xEV-DV is not even slideware yet.
8.) Wireless is a standards world. Little gets implemented that is not standardized. Full commercialization begins when standards complete. It would appear that the "first to market" advantage possibly goes to W-CDMA, and that cdma2000 lags W-CDMA in standardization, any publicity claims (related to interim steps) aside.
9.) While cdma2000 focused heavily on forwards and backwards compatibility with cdmaOne (IS-95A/IS-95B) using the "evolved" ANSI-41 core, UMTS focused on forwards and backwards compatibility with GSM using the "evolved" GSM/MAP core. GPRS and EDGE standards have evolved concurrently with UMTS standards and have the same focus.
10.) "Harmonization" is not "Convergence". Harmonization provides some degree of interoperability. It does not provide interoperability to the degree that convergence does. Intermodal and internetwork interoperability become important factors in the new global third generation game.
11.) Multimode - multiband handsets are a reality but the more modes and the more bands you add the more complex the handset becomes. Eventually Software defined radio (SDR) will solve this issue, but that is tomorrow, not today.
12.) GSM carriers (the majority of worldwide carriers) are HIGHLY UNLIKELY to choose any technology whose standards are not under the control of their own standards body (formerly ETSI and now 3GPP). Their massive strength and success is due to their robust, evolving standard and their spirit of coopetition (rather than competition) with there fellows that they gain from participation in GSMA.
13.) GSM carriers of necessity need the assurance that GSM, GPRS (and maybe EDGE) are fully interoperable. They will in some cases operate all in the same geography or market. The will operate, in many cases cooperatively with their competitor while building out as they did in the early stages of GSM buildout. This is in sharp contrast to PrimeCo which didn't interoperate with BAM and AT which didn't interoperate with Sprint PCS. etc (and Verizon and Sprint PCS still don't). Europeans still scratch their heads in wonderment over this.
14.) IS-136 TDMA carriers are unlikely to choose any technology whose standards are not under the control of their own standards body (formerly TIA through UWCC and now 3GPP GERAN). It is of course possible to see a "technology flip" and this may apply to Bell South International who is a CDG member and operates both TDMA and CDMA in LA. The association of the UWCC with GSMA is recent but adds to the strength of GSMA.
15.) Korea specific (see point #2) - Korean carriers have been experimenting with and trialing W-CDMA for some time. Korea submitted 2 asynchronous W-CDMA proposals (similar to DoCoMo's to ETSI) when the air-interface for UMTS was being decided. It appears that they feel that operating CDMA using a 5 MHz carrier is a better way to optimize CDMA in 2GHz spectrum than using a 1.25 GHz carrier. No, we can't rule out a pending DoCoMo investment in SKT as a decision factor, but it is deeper than this.
16.) Voice and Data Roaming are major decision factors in selection of 3G technology. They evidently are bigger factors than some people realized. Perhaps Roaming should be considered a HUGE decision point. Roaming has long been a focus of GSMA. This applies to voice roaming, SMS roaming (15 BILLION GSM SMS text messages will be transmitted this month, and growth - much of it due to roaming - is exponential), and now IP roaming. The GSMA emphasis on roaming reflects in the UMTS standards and probably to a much greater degree than in the cdma2000 standards. It also is reflected in costly apparatus that has been implemented to make it practical. CDMA operators and CDG (in the past) have placed considerably less emphasis on roaming.
17. Scale (relative to roaming and interoperability) - There are over 400 GSMA and UWCC carriers in the world operating in 161 countries. There are 31 CDMA carriers in the world.
18.) Vendor financing frequently plays a major role in technology selection. It certainly did when PrimeCo and Sprint PCS chose cdmaOne over GSM-1900 in the summer of 1995. It is definitely playing a role here. We are hearing of "deals" where vendors are financing as much as 125% to 300% of the cost of the network buildout. In addition vendor financing often is accompanied by performance guarantees and overall reduces risk. Vendor financing may be more readily available for W-CDMA than cdma2000. I suspect it is.
19.) Economies of scale - Given the number of GSM operators in the world (and including those that are becoming GSM operators by virtue of UMTS technology selection) economies of scale favor UMTS over cdma2000. While there may be some cost advantages to cdma2000 (or 1xEV-DV when it becomes a reality) these are probably mitigated to a great degree by the manufacturing economies that will accrue to UMTS. In addition this means that at least initially we are likely to see a larger variety of W-CDMA devices than 1x/HDR devices.
20.) Overall Cost - The initial cost of new networks or network upgrades for 3G (CAPEX) are a very long term investment. They are a significant component of owning and operating a network but by no means the only one. There is licensing, and marketing, and distribution, and R&D, and services development, and maintenance, etc. As for replacing existing (legacy) componentry with new as opposed to upgrading, replacing with new is sometimes desirable and more cost efficient in the long haul than upgrading. It remains to be proven whether cdma2000 is really more cost efficient and cost effective) than UMTS, even if it requires higher CAPEX initially.
21.) Risk Mitigation - cdma2000 operating in 1.25 MHz spectrum is commercially proven technology pioneered by QUALCOMM. There is definitely more risk associated with deploying W-CDMA than there is 1xMC (and perhaps even commercially unproven HDR). Risk is somewhat mitigated by the fact that virtually every wireless equipment manufacturer in the world is fully committed to making it work, and making it work well. Venor financing also mitigates risk for a carrier.
22.) W-CDMA is an open (comittee based) standard. cdma2000 is perceived to be (is) an open but proprietary standard. Carriers prefer open standards to proprietary standards (open or closed) by a wide margin.
My opinions are above. Enough of me. Recently Ira Brodsky (industry consultant and long time CDMA champion) said:
"cdma2000 has major advantages over W-CDMA. cdma2000 will deliver full IMT-2000 capabilities (speeds up to 2 Mb/s) in one-third as much spectrum. Commercial cdma2000 service will be introduced first most likely early 2001. And cdma2000 is backward compatible with the cdmaOne technology already used by nearly 70 million subscribers.
W-CDMA, the preferred solution in markets dominated by GSM, will also be successful. W-CDMA will be first deployed in new spectrum allocated for 3G services; then it will slowly replace GSM and U.S. TDMA in existing spectrum. However, W-CDMA is likely to take longer to develop and commercialize than proponents claim."
Now with all that said and done, my good friend Cha2 will attest to the fact that since I have held QCOM, I have maintained that UMTS would be the technology choice of at least 70% of the carriers in the world and it appears to be turning out that it will be the technology choice of at least 85% of the carriers in the world.
So does this make QCOM a poor investment. My answer is a resounding NO.
Whether cdma2000 (1xMC/1xEV) or UMTS UTRA we are talking CDMA,and QCOM gets paid on CDMA.
Andrew Seybold recently said:
"Consider this. All 3G systems are ultimately based on CDMA technology. This means that even those systems that now use two competing standards - GSM and TDMA - will eventually turn to CDMA when they move to 3G systems.
Nor is Qualcomm a loser yet in Japan. It was recently granted a patent there that is equivalent to one of the first patents it received in the U.S. more than ten years ago. What it says is that Qualcomm is entitled to receive royalty income from any CDMA-based network architecture including Wideband CDMA and UMTS-2000, the European version of W-CDMA. Qualcomm has already signed agreements with a number of companies, including Samsung, Lucent, Ericsson and Sony, that have agreed to pay it royalties regardless of what version of CDMA they use in their products."
I believe, Andrew on this. I believe, Irwin on this. I would believe it without them.
Now as for 3rd Generation takeup Brodsky also recently stated:
"CDMA Will Unleash the Internet. Next Generation Wireless Devices Will Number More Than 500 Million by 2005."
Yes, I would prefer to see more cdma2000. I am however anxious to getting the 3G3 CDMA SHOW on the road. Flavor be damned.
500 million Next Generation Wireless Devices of any CDMA flavor sounds pretty good to me.
- Eric - |
