Wednesday, January 26, 2011

Mobile Network 101 Finale

Back to the question I asked at the beginning, why are the pre-4th generation iPhones no better off than a cheap Nokia in New Zealand?

The answer is simple: When the iPhone was first sold in NZ, Vodafone was the only carrier with a GSM service so Apple had no alternatives. However, iPhone was designed for AT&T's 850MHz 3G frequency instead of the European 900MHz that Vodafone uses. While the older iPhone models are able to get 3G service in urban areas where 2100MHz masts are the mainstream, they are out of luck in less populated places as their phones will only get 2G/GSM signal. Ironically XT had the right frequency nationwide, however Telecom opted to offer iPhones through its subsidiary Gen-i only. 

The same issue is not limited to one brand. Motorola A855 or Sholes came in several versions that are nearly identical except for the radio module. The GSM/WCDMA version is called the MILESTONE, which is further divided into European (900/2100), Commonwealth (850/900/2100) and North American (850/1900) flavours, causing much confusion when users move across national boundaries; and the DROID line is CDMA2000 only and without a card slot, making them effectively unusable outside the US, unless you have a friend of a friend to unofficially register your phone with a non-US carrier.

Sometimes it is harder to determine if a certain phone is XT-compatible since 850 and 900 versions may be offered under the same model/part number. A friend working with parallel imported handsets often had to literally sail out, with his newly arrived stock, to a spot in Hauraki Gulf that he knew has only 850MHz coverage.It may sound ridiculous but there are no other ways to tell.

Apple has certainly learned something and the iPhone 4 is given a pent-band 3G baseband; theoretically it should work in any place with some form of WCDMA service.(The FCC documents included a 800MHz band however Apple choose not advertise on this, probably because there are not many 800MHz networks out there) iPhone for CDMA2000 is also due to be released shortly. Motorola has also recently introduced a range of Global phones with both GSM, WCDMA and CDMA2000 hardware, however these phones are programmed with a new type of SIM lock not to work with GSM carriers in the USA while they work without restrictions elsewhere.

Nevertheless, Blackberry has been selling truly global phones for many years so it must have never been a major technological hurdle to combine several radios in one device. The true motivation lies in the lucrative practice of carrier subsidy.

My conclusions from the history of mobile telephony

  • Like all other forms of infrastructure such as electricity and railway, mobile networks are costly to deploy and convenience decisions often lead to many headaches later. Example: Telecom took the hard pill in giving up CDMA2000 altogether,however many telcos elsewhere are still spending billions of dollars every year to expand their present CDMA2000 network and upgrading them to EV-DO Ver.B for commercial reasons with the full knowledge that the system is minimally compatible with successor standards and in a way, already obsolete.
  • Homogeneity created through monopoly is bad, too much diversity is worse.
    • Conforming to the general patterns in industrialisation, latecomers often have considerable advantage since they are not already committed to maintain legacy support. There are many examples: Japan, once they got over the general ineptitude for much of the 2G era, successfully developed a global standard and currently has several of the best and most profitable 3G netoworks world-wide.
    • Theoretically superior technology does not always lead to better results. Example: Telecom's XT is definitely faster than Vodafone's older UMTS stations, however XT is still struggling to catch up with its own precedents in terms of coverage and reliability, the two essential criteria of any good mobile service that appears to have been overlooked in the last 10 years. 
    • Lysenko may have been dead for several decades, his spirit still lives on as purely scientific matters are often swayed by political and/or ideological influences. The same farce is ongoing with the entire WiMAX vs. LTE debate.
    • Tanenbaum noted that public interest (and investment) in scientific advancements usually arrive in waves, hence the evolution is more or less stochastic. Example: 3G was almost killed off following the .com bubble, and it is still struggling to return a profit amid the recent recession.
    A lot of poorly developed science remained alive and well through pure luck, while the most ingenious inventions may easily slip into oblivion if it was born around the wrong time.

    By the way, his book Computer Networks is an essential read for those with further interests in this area.