An apocryphal story was reported recently of an iPhone user in New York who wanted to return it and have it repaired because of the large number of voice calls which were failing. On investigation, the store found that there was a 22% call drop rate. However, this wasn't deemed to be particularly unusual for that area and the customer was sent on his way without any change. Wouldn't it have made him a happier customer if he'd been given (or even just able to buy) a femtocell to solve these issues when indoors?
22% call drop rate identified
Gizmodo reported what the customer had been told, even publishing the written report from the Apple Genius diagnostic team. They had said that typical drop rates are nearer 30% in the city, and that 22% was "excellent".
This drop rate was arguably higher because the customer manually terminated calls when they started to fail, reducing the possibility they might recover later.
For those of us in many other parts of the world, we'd find this level of performance quite shocking.
Gizmodo notes that a femtocell (ATT's 3G Microcell) might help a lot, but doesn't think the customer should have to pay for it.
Why is it so bad?
Here's some ideas (I don't have any specific inside information here).
The first question to ask is whether the problem is coverage (no signal/no bars shown on the phone - common indoors at home) or capacity (5 bars signal on the phone but interference is so high the system can't take any more traffic). Ideally, both need more cellsites to solve the problem.
New York is a city with lots of skyscrapers and lots of users. RF transmission patterns must be a nightmare to design, with competing effects from users in buildings and on the street. ATT's network operates mainly at 1900MHz - at these frequencies the signal doesn't travel so far, doesn't penetrate well inside buildings. This can be great when trying to add extra capacity through extra cellsites, but makes it more difficult where the signal needs to be higher power to reach inside buildings - this causes higher interference in adjacent cells and reduce overall network performance.
Other parts of the world also use the GSM standard, but usually at 900MHz frequency. Some operators are at 1800MHz only, but even these have often manage to trade or been allocated some spectrum in the 900MHz band (e.g. all 4 GSM operators have been allocated frequencies in both bands). Where 3G is used, at 2100MHz, users can always fallback on the 900MHz GSM service where signal quality is poor or unworkable.
The recent availability of some more lower frequency spectrum at 700MHz and 850MHz which is being used for 4G network rollout will help fill in these poor coverage gaps.
Would offloading the traffic help?
ATT clearly think so. They bought Wayport for $275M last year and now have some 20,000 WiFi hotspots which their customers can use. This week they announced that any customer with a Microsoft Windows Mobile phone could access any these free of charge. They've said that at least 30% of iPhone data traffic is offloaded through WiFi today.
This not only benefits the user who should receive faster data service because they are sharing a lot of capacity with only a few others, and are close enough for the signal to be able to achieve high throughput. It also benefits other users outside who no longer have to share the congested traffic channels.
It's a bit like giving free metro passes to every citizen and seeing the roads free up in city centres.
Femtocells would also achieve the same effect through network offload with the added advantage of being easier to access and to control because it uses licensed spectrum.
Traffic offload to WiFi and Femtocells has to be part of an operators long term plan to deal with the data capacity crunch and avoid the bottleneck/congestion problem.
To achieve this, dissatisfied customers like the example above should be given a free femtocell (perhaps on contract or as part of a package) which would both make them happy and relieve the outdoor macro network so that it can serve the wider userbase more effectively.
Arguably, it's also a cheaper way to rollout substantial additional network capacity than installing more outdoor cellsites.