Wednesday, April 6, 2011

Mobile Phone Antennas and Networks

We all remember the so called 'Antennagate' where the iPhone 4 loses coverage due to the way its held. As can be seen from the above picture, there are a lot of antennas already in the phones and yes they are on the increase with LTE and other technologies being added all the time.

Apple admitted the fault and claimed to have fixed the problem but its well known in technical circles that the fix is more of a software hack which doesn't really fix the problem just pretends to fix it. That is why the networks dread it and you can find awful lot of information on the web about the problems.

In a recent Cambridge Wireless event, I heard an interesting talk from Trevor Gill of Vodafone and one of the slides that caught my attention was the impact of these poorly designed phones on the network. The slide is embedded below.

It is estimated that the RF performance of iPhone4 is around 6dB worse than most other 3G phones. What this means is that you may be getting 4 bars of reception on your other phone where iPhone4 may be having only 1 or 2 bars or reception. So if the reception is poor with 1 or 2 bars, iPhone4 may have no reception at all.

To fix this problem, either the networks can increase the number of base stations to double the existing amount which is a huge cost to the networks and extra radiation or the phones can fix it themseles by having an extra antenna. In fact as the slide says, extra antenna on each phone would translate to increase in network capacity by 20-40%, cell area by 30% and cell edge throughput by 40-75%.

One final thing that I want to mention is that testing (RF, RRM, Conformance, etc.) are mandated by the networks for most phones but they overlook the testing procedure for phones like iPhone. What this means is that they do get a lot more new customers but they get new sets of problems. If these problems are not handled well, the impression they give is that the particular network is rubbish. Another thing is that the devices use a certain build/prototype for testing but the one that they release may contain other patches that can cause chaos. One such problem was Fast Dormancy problem that I have blogged about here.

Hopefully the networks will be a bit more careful and will put quality before quantity in future.

2 comments:

Steve Crowley said...

Another idea is to let users have inefficient phones if they want, but charge them more for the additional network resources required.

Rummy said...

I read this post in April. During discussion with colleagues about the WCDMA preferred option in Android OS this post flashed back to me.
As mobile communication moving toward 2/3/4G mixed network, the RAT preferred option implemented on UE side is somehow tricky and the impact to the network is unknown. The need of this kind of feature is often triggered by some certain scenario happened in some certain location and it's operator dependent. Usually, for non-tech-related reasons, it's difficult for engineers like me to get information from operators to know if this kind of features are designed well or ill. I hope smart UE designers in the world not to do tricky features that might cause undesirable behaviors. Especially to UE designers who behave well to pass regulatory tests in lab but magic their devices in the real field, please don't.