How the Chinese screwed up their 3G mobile phone networks

In April, a Chinese government minister proudly announced that there are now 152m people using 3G mobile phone services in China.

An ongoing story in the Apple world is that the world’s largest mobile phone company, China Mobile (which has about 650 million customers and 70% of the mainland Chinese market), does not offer the iPhone to its customers. There has been much speculation as to when Apple and China Mobile will “do a deal” in order to offer the iPhone to its customers, many stories written by financial analysts (many working for large equities houses) expressing their puzzlement about the lack of such a deal and pressuring the two companies to get on with it. There has been much discussion as to what the business reasons for the lack of such a deal might be – possibly the two companies are at odds about shares of revenues when they sell Apps and music and movies to iPhones over their networks. Such arguments have long characterised arguments between handset vendors and mobile networks elsewhere in the world, although generally Apple has won the argument outside China.

This seems puzzling, however. There is a great deal of money to be made, right now, and this money is simply not being made by anyone at the moment. It requires parties to be very stupid indeed to concede that two parties will have 100% of nothing, rather than undecided shares of many billions of dollars, which is apparently happening here. Neither Apple nor China Mobile are stupid.

This isn’t it. There is more going on here. Or possibly less.

Last week, another piece of information came out of China. The real number of 3G users in China was in fact only half this – between 75 million and 80 million. China Mobile had “mistakenly” provided the whole number of devices sold using its TD-SCDMA technology rather than the number using 3G data services. As well as using this technology for 3G data services, China Mobile also uses it for cordless (but fixed) telephone services. The vast bulk of devices sold had been very simple essentially fixed telephones that could just be used for simple voice and text services.

So how many 3G customers does China Mobile have? Well, there are two other mobile networks in China. China Unicom operates a 3G network using W-CDMA (people who refer to “UMTS” or “HS(D)PA” are generally also talking about this technology, although the terms actually mean slightly different things), the technology used exclusively in Europe, by AT&T and T-Mobile in the US, by NTT in Japan, and by the majority of carriers elsewhere in the world. China Telecom operates a 3G network using the CDMA2000 (often referred to simply as “CDMA” in the US), used by Verizon and Sprint in the US, KDDI in Japan, and by many carriers elsewhere in the world, although far fewer than the number that use W-CDMA. China Unicom apparently had 51 million 3G customers in April. China Telecom claimed 45.56 million.

China Telecom’s number is easier to exaggerate than is China Unicom’s number, as the CDMA2000 technology was developed as a gradual evolution of the (2G) CDMAOne technology, and where you draw the line between “2G” and “3G” is somewhat arbitrary for this technology. W-CDMA was an all new air interface compared to its 2G predecessor, so (if you are being at least reasonably honest, which is never guaranteed in China) it is harder to fudge the numbers than it is for CDMA2000. Assuming that the 51 million number for China Unicom is relatively sound, and bearing in mind that China Telecom has around half the overall market share of China Unicom, a fair estimate might be that China Telecom really has around 25 million 3G customers, assuming that “3G” means people achieving similar data speeds to what is being achieved on China Unicom’s network. (China Telecom clearly do have a significant number of 3G customers. 3G devices bearing their brand were readily available and being sold aggressively in the electronics markets of Beijing when I visited last November, as were those of China Unicom. China Mobile, much less so). So that would mean that the two smaller carriers, China Unicom and China Telecom, have around 75 million 3G customers in all. Which would suggest that China Mobile, the largest mobile phone operator in the world, has, as a first order estimate, none.

This might help explain why they are unable to offer the iPhone.

So, it appears that the largest Mobile carrier in the world, in a country perceived as being rapidly advancing technologically, may have no 3G customers. Why is this?

To answer this question, we need to go back around 20 years. In the early 1990s, the European Union (and the European Telecommunications Standards Institute or ETSI) decided that a single technological standard should be mandated for digital (ie 2G) mobile phones in Europe, to avoid the national fragmentation that had existed for analogue (1G) services in Europe. This had two features that turned out to be particularly useful. Firstly the phone number and account was separated from the phone handset, instead being associated with a SIM (essentially a smartcard). Secondly, easy international roaming was mandated, and a phone used in one country using the standard would still work when taken to another country using the standard. (In my mind, certain aspects of how resultant choice of networks and billing were handled were botched, and this is a large cause of why international roaming charges remain so horrendous today, but that is a different story). The actual technology of GSM was in certain ways quite primitive, but the resulting system worked very well, at least for voice calls.

In any event, the resultant GSM standard was adopted in Europe by mandate, gaining critical mass. Because GSM hardware was readily available and the standard was well defined, the GSM standard became the most widely used in most of the rest of the world, other than in the Americas, Japan, and Korea. Economies of scale have meant that other standards have gradually lost ground to it and its successors ever since.

“The rest of the world” included China. Two Chinese mobile networks were set up, both of which used GSM, and both of which belonged to the Chinese government. China Mobile was the favoured company, which became dominant, and China Unicom was the secondary competitor.

However, the (principally European) companies that had invented GSM had a heavy patent portfolio, and demanded significant royalties be paid for use of GSM and GSM related technologies. China resented this. This was the situation in around 1995.

In the second half of the 1990s, something called the IMT-2000 process took place within the International Telecommunications Union. This was an attempt to come up with a single international standard for third generation mobile phones. A system called W-CDMA, that had initially been invented by NTT of Japan, but which was championed by Nokia and Ericsson was adopted by the European ETSI (ahead of an alternative standard invented by Siemens) as the European standard that they put forward to be the international standard. Qualcomm of the US put forward an alternative standard that was an evolution of their existing (2G) CDMAOne standard, and this eventually became CDMA2000. No agreement could be reached between all these parties, which is why there are two common 3G standards in the world today, the aforementioned W-CDMA and CDMA2000.

China, as I said, resented the royalty payments that were being demanded by either of these parties, and decided that they would not adopt either of these standards, but instead would develop an indigenous Chinese standard as an alternative. In around 1999 and 2000, Chinese bodies started issuing press releases stating that China would be adopting neither W-CDMA or CDMA2000, and that an indigenous Chinese standard called TD-SCDMA would be used for 3G mobile phones in China.

China did not have the technological skill to develop a 3G mobile phone standard in 2000. So where did they get the idea that they could and would develop their own? Go to the Wikipedia page for TD-SCDMA, and you find this interesting paragraph

TD-SCDMA was developed in the People’s Republic of China by the Chinese Academy of Telecommunications Technology (CATT), Datang Telecom, and Siemens AG in an attempt to avoid dependence on Western technology. This is likely primarily for practical reasons, since other 3G formats require the payment of patent fees to a large number of Western patent holders.

The most interesting words in this paragraph are Siemens AG. What appears to have happened is that having lost the battle to provide the technology for the European standard, Siemens of Germany then attempted to sell it to the Chinese, on the basis that they would not have to pay the royalties that they would have to pay if they adopted the winning standard. The Chinese apparently were delighted by this suggestion, and started almost immediately issuing press releases about how they were developing their own standard.

People who remember the telecommunications industry of a little over a decade ago will also remember that the road to 3G in Europe using the adopted W-CDMA standard was a fairly bumpy one. Huge amounts of money was spent by operators on 3G licences, and then the tech bubble collapsed, the technology turned out to not work very well, and in the mid 2000s we were still using 2G mobile phones happily enough. (Seriously, in 2005 you were probably using a Motorola RAZR. Think how primitive that is compared to what you have now). It was only with the spread of 3G broadband dongles and the iPhone that 3G really got going and the technology really started working well.

This was with all the capital and expertise of European and American firms. In China, engineers with many fewer resources (including German resources, but still fewer) were working with a technology that had not been adopted in Europe and Japan due to being an inferior technology in the first place. The Chinese had announced to the whole world that they were developing their own 3G technology rather than use foreign technology. There were only two things wrong with this. The technology had in fact been invented by the Germans. Also, it didn’t work.

When I tell this story to Western expatriates in China, they start laughing and making gestures that consist of (amongst other things) pointing at their faces. Losing face is very, very bad in Chinese culture. We had a great many Chinese politicians and bureaucrats, engineers, telecoms executives with a lot of face caught up in this. The sensible thing in about 2006 would have been to have dropped the whole idea of the supposedly indigenous Chinese 3G standard, but the idea of face would not allow this.

However, the 2008 Beijing Olympics were coming up. China was trying to sell itself as a modern and advanced country. The Chinese were horrified by the idea that vast numbers of the international elite would come to Beijing for the Olympics, discover that there were no 3G networks, and conclude that China was a poor, backwards place, and the Chinese elite are even more frightened of this than they are of losing face of mobile phone standards, because deep down they are frightened that it might actually be true. (The whole point of the 2008 Olympics was to demonstrate that it was not).

So, it was decided that China *must* have 3G networks in time for the 2008 Olympics. But face must not be lost. So China Mobile, the favoured and dominant (state owned) mobile carrier was ordered to build a TD-SCDMA network at once, whether or not it worked. As the most powerful operator, it should use the Chinese indigenous standard.

As it was necessary that sophisticated foreigners who came for the 2008 Olympics should be able to have their phones roam to 3G networks, somebody else had to be operating a W-CDMA network by the time of the 2008 Olympics. The obvious choice was that China Unicom, the second, less favoured, state owned mobile operator should build an internationally compatible W-CDMA network in time for the Olympics.

There was only one problem with this. China Unicom had a 3G network already.

Er, what?

Huh?

As in many countries, in China, it had been seen as desirable for there to be competition between different phone companies. Fixed line operators were encouraged to get into mobile, and mobile operators were encouraged to get into fixed line. Competition in such things was heavily encouraged from about 1990-2005, but this has fallen by the wayside a bit in the years since, as mobile has simply become dominant).

The trouble with competition in fixed line is that the legacy monopoly carrier generally owns the copper wire from the exchange to the home, so a competitive local phone company must either lease that copper wire from the legacy carrier, or find some alternative way of getting a connection to the home. Various things have been tried: using cable television cables, electricity wires, water pipes, microwaves, or radio signals of various kinds. Qualcomm, the American company that had invented CDMAOne/CDMA2000, was quite successful at advertising its CDMAOne/CDMA2000 technology as a solution for companies that wanted to offer competitive fixed line services. (Wireless Local Loop, or WLL). The idea was that you would install a phone in your home that would connect wirelessly to a local exchange, and that you would only use this phone in and near your home. If you went away from your home, the phone would not work with other exchanges, and the phone could not handle handover from one exchange to another if you were on the move.

Of course, CDMAOne/CDMA2000 was in fact a full cellular mobile system that was perfectly capable of these things, but if it was purchased as a WLL solution they would be switched off. (They could, of course, be switched on at any time).

In order to compete with existing fixed line carriers, China Unicom had been given a licence to use some spectrum for Wireless Local Loop. They rolled out a network using Qualcomm’s CDMA2000 system, and as far as I can tell, they switched everything on. If you had a handset and went a long way from your home, your phone moved from network to network. Although China Unicom did not have a licence to build a 3G mobile network, they built one anyway. And the powers that be in China either failed to see this or were bribed not to.

But, 2008 Olympics. It was necessary to build a W-CDMA network in order to prevent those influential foreigners from coming to China and laughing at how backwards China was. (Actually, the foreigners had their phones switched off because roaming charges were too high, but that is a different – although not entirely unrelated – story). Also, China Mobile, the largest and politically best connected mobile phone company in the world, was pissed off enough already by the fact that they had been required to build a 3G mobile network using a standard that barely worked. If China Unicom was allowed (or required) to built a W-CDMA 3G network in addition to the CDMA2000 3G network that they already had, this would be dangerous from a competition perspective, as China Mobile could be overwhelmed by the superiority of the technology of their main competitor.

So, the Chinese government had to hobble China Unicom. Firstly there is no number portability in China to this day. In virtually every other country in the world, a customer may change mobile networks without changing their phone number. This is a huge deal. People are very reluctant to change their phone numbers. Churn rates (ie the percentage of people who change carriers per year) are much higher in markets where number portability exists.

Secondly, the Chinese government decided that more competition was in order. Fixed operators should compete with mobile operators to a greater extent than had been the case previously. The Chinese telecommunications business should be “restructured”. The (state owned, as always) purely fixed operator China Telecom should also operate mobile services. It would be a huge obstacle for this company to build a mobile network from scratch, but because China Unicom owned one more mobile network than it needed, Unicom would be required to sell off its CDMA2000 network to China Telecom. That way, there would be three operators, each with one 3G network, each using a different technical standard. China Mobile would have TD-SCDMA, China Unicom would have W-CDMA, and China Telecom would have CDMA2000. Each of China Mobile’s competitors would have a superior network to its own, but each would be competing with one another as well as China Mobile, and each would be hobbled in competition with China Mobile in various ways.

However, face needed still to be saved. Evidence that China’s indigenous TD-SCDMA technology was a success needed to exist. It seems that China Mobile (who were of course being encouraged to compete with fixed line operators) discovered that although TD-SCDMA was not especially useful for 3G mobile services, it was good enough for those Wireless Local Loop services that mobile carriers were being encouraged to provide. So whereas China Unicom had previously managed to provide 3G mobile services by claiming that they were really Wireless Local Loop (for which they had a licence), China Mobile has been able to provide Wireless Local Loop services with a barely adequate technology and claim that they are really 3G, because that is what the Chinese government wants them to claim. The hardware has been cheap and practically given away, and its (possible) adoption has given the China Mobile the ability to claim that their number of 3G customers is greater than it really is.

Except that they have just been found out. I wonder how that happened?

Perhaps we are at a moment in history when everyone, everywhere, is being found out. If so, I am scared.

Apple has not been willing to produce a version of the iPhone for China Mobile’s network. This is because they cannot make a version of the iPhone that works on that network well enough for Apple to be willing to put their name on it. It really is that simple.

To be fair, this will go away before too long. For 4G, China is going to use a variant of the LTE standard being used throughout the rest of the world. The TDD variant used in China is going to be used elsewhere as well, and the FDD variant that is presently being used elsewhere will likely be used in China as well. Handsets will likely handle both. There may be some insistence in China that they are still using a compatible indigenous standard, but they are smart enough to not repeat the same mess again. Everyone in the world looks set to adopt compatible standards this time. But oh boy, it has been painful to get to this point. (The number of different frequency bands that is likely to be used will present handset manufacturers with a little bit of a challenge, but I suspect they will meet this).

Meanwhile, India has half a dozen excellent W-CDMA 3G networks owned by half a dozen different companies, many of them non-Indian. But that is another story – one I will tell soon.