Does Apple Bite?
In the beginning (and remember this is just a parable): A cool manufacturer developed a unique device called Handset X. The idea was to launch it originally with one operator, create a big splash, and grow from there.
The Handset X folks launched the 2G version of that handset with one operator, tied with a bunch of unique apps and utility. Getting the first Handset X right required a huge staff, significant investment, tons of time, and deep coordination between the Handset X folks and the launch operator. After a hugely successful launch of Handset X with one operator, working through a broad array of technical and market issues, the 2G version of Handset X was launched with a limited set of other wireless operators globally. Handset X’s customers loved the device, and eagerly awaited the 3G sequel.
So its time for a 3G version of Handset X: Making 2G versions of devices, especially in the GSM/GPRS/EDGE world is not that challenging. Not easy, but not like starting space tourism. There have been billions of phones made. There are over 2 billion plus users. There are hundreds and hundreds of GSM operators in almost every country around the globe. The service and application platforms are relatively mature, mainly voice, SMS, WAP, simple browsing, and simple apps. The radio and antenna side in the handset are also well established, the vendor base vast for sourcing manufacturing parts, and radio coverage of the mature GSM networks for these devices is typically very good to phenomenal.
So, the 3G version of Handset X gets launched with much fanfare. Over the weeks, there is a lot of noise in the blogosphere and press about bad battery life and 3G connection issues. What could this mean about Handset X? Well, 3G for Handset X means that it has a 3G radio, which in their case means having both WCDMA and HSDPA radio in the device, as well as multiple radio bands and antennas.
Typically in a 3G system, if battery life is bad, that means that the device, Handset X in this case, could be using too much output power to speak with the antenna at the operator’s cell site. Handset X is having to "shout" and just generally work too hard to be heard by the cell site, and it takes more calories to shout than to whisper. In Handset X’s case, this can be a result of many issues: the chipset, the antenna(s), the RF filters, the software in the phone that runs the radios, and a myriad of other potential issues with handset design.
Or, if not handset design, it could be in the testing of the handset. Maybe the testing protocols are immature, or maybe production has ramped so quickly that testing is variable across different manufacturing lines. Beyond factory testing, how well does Handset X “interoperate” with various wireless operators’ infrastructure, as that infrastructure could be from a range of vendors (i.e. Nokia-Siemens, Ericsson, Alcatel-Lucent, ZTE, etc.)?
In Handset X’s case, they just did not have much experience developing 3G handsets, and without tearing down the Handset X device, I’m just going to assume they made a bunch of rookie mistakes, the types of mistakes that would have been made by some of the top global handset companies four or five years ago that have been long since solved. Remember, in this example, the company building Handset X is building their first 3G phone, unlike other companies that have been at this for years, and have made all the mistakes there are to make (and are still working on issues like real usability!)
The cell site saga
The story gets worse for Handset X: Next, one must look at the infrastructure and the cell sites. When an operator deploys 3G they have to figure out where to put their cell sites. This is a complex and expensive process, as the cell sites typically cost 80 percent or more of the cost of any wireless network. So, operators will not overbuild with too many cell sites, especially since nobody wants new sites in his or her backyard, and it costs a fortune and can take years just to get through the legal approval and zoning process for a new cell site.
So, imagine a city or suburb map, covered with concentric overlapping circles. In a simplistic view, this is called a "network plan." It's simplistic, as an open flat sandy desert may be easy to cover, but radio waves do weird things when they have to bounce off buildings, go through trees, more importantly, try to penetrate buildings that range anywhere from wood to concrete and iron.
So, there are a bunch of phones by a bunch of manufacturers on the operator’s network. They are all talking to the cell site. Then an interesting thing starts to happen: As a 3G cell site starts to get under load, the range (or cell radius from the cell site tower) starts to shrink. And if the cell site shrinks, guess what happens? The handsets on that network, including Handset X, will need to talk louder to be heard -- using more power, and if a handset such as Handset X has design or technical problems, that will drain battery life even faster than before, and it will potentially have more connection issues! The more people using the network, the worse it will be for Handset X!
A funny thing with this variability -- to look at things simplistically -- wireless coverage is "circle" around the cell site. If a user lives, works, and plays near the "center of the circle" they may just be fine, and many early Handset X users reported that they had no troubles at all. But if someone lives, works, and plays near "fringe" coverage areas, with a Handset X disadvantaged by any of the issues above, it could be a problem.
How could a phone have been approved and fielded with these issues? Well, even if Handset X’s factory side was nailed down, maybe they did not do enough field testing on the device. Maybe the field testing was mainly done in lab environments, or maybe it was done in an area where the launch operator had overbuilt 3G coverage. Maybe it was an area that was suburban, tilt-up business construction, which is easier for radio waves to penetrate. Maybe it was Silicon Valley.
Maybe the pressures to launch trumped or buried the concerns of the Handset X designers and field engineers that were doing the testing. It would not be the first time that’s happened in the history of tech products.
So, we’ve covered Handset X’s first challenges, so where could it get worse? Everywhere.
Why, because coverage, morphology (i.e. the types of terrain that the wireless needs to cover), and in-building penetration, as well as the types of devices and "start" network load, are challenges all over the globe. The types of networks that Handset X will need to work on will be more challenging than their initial local launch network.
In a lot of places around the world, 3G networks are still maturing from a coverage/in-building penetration perspective and often rely on the increasing RF performance of today’s “Tier 1” handset designers. Plus, for many of those networks, there is great adoption of mobile broadband in laptops, either embedded or via USB dongles. This category of devices is taking off, will continue to accelerate, but last time I checked, a one-pound/500 gram laptop battery has a bit more fudge room for power inefficiency than a 120 gram Handset X. There’s not much wiggle room to go back in time for less efficient handsets.
Now in this case, we’re not sure if it’s the phone software, the chips, or the RF front end, testing, interoperability -- or a combination of these and others. But the bottom line is that it won’t all be fixed with a software patch! So, if the folks that make Handset X are smart, they will cut their losses and make sure that the next hardware iteration of Handset X knocks off most of these problems!
Page 4: Coming soon: Apple WWAN laptops?