No Service (what a surprise)

Question

I have been without phone service for a day and a half. This is the second time in the last two weeks. This time ATT swears nothing is wrong with their network.
Any suggestions? I have gone through all the steps in the troubleshooting guide except restore. I can't do that because my powerbook is currently at the Apple store getting the hard drive replaced (it's been a bad week)
Interestingly enough, the last call I was able to make, I made from the Apple store!
Outside of not functioning as a phone, this is really a great phone!
Any help is appreciated

Answer 1

I must say that I'd be surprised to find a SIM replacement much affected a variable signal problem (though in fact I have had that very experience years ago in the UK so I know it can be a valid fix), but for the OP here where signal appears to have been lost completely (or nearly so) and where AT&T declare the network in that area is functioning, the SIM is the first thing I'd consider culpable.
The issue regarding conflicting towers is certainly true however. It is rare that a location with any cell coverage is serviced by only a single transmitter on a single tower, and in many cases it's possible to find multiple towers all within range, operating transceivers on more than one band the cellphone can lock to. I'd assumed that some of the signal variation I see is at least the result of metering two bands from towers in range, where one has better penetration into the building and thus gives higher signal. In those situations, inductance caused by the proximity of the body and many other small factors could cause the phone to vacillate between bands/towers.
Of course, much is speculative since neither AT&T nor Apple are all that open to describing the details of their service/equipment/devices, but I would be surprised if this sort of issue were not something that could be addressed in software. Indeed, I'm not sure from my experience with signal strength metering in the last 36 hours that some attempt hasn't already been made since I get far less dramatic variations in signal now that I did before the 1.0.2 update, and that is despite crossing a number of cell boundaries. Even so, it isn't 'fixed' if there has been an attempt.
One way or another though, it's too easy to fall into the trap of assumption. Cellphones are radios, very lower powered ones, and as such the expectation of solid signal is almost inevitably going to lead to disappointment with a cell structure which is relatively sparse and where tower placement is often determined by cost and convenience rather than service need. We also don't know for sure how much variation in metering we see is actual variation of signal rather than variation of metering, and whether the metered signal is reflective of viability of service. I say that because I have noted that if I see a drop in signal while on a call, the quality of the call seems totally unaffected, while if I start a call on very poor signal, it seems to suffer audio artifacts. To some extent I wonder if the meter reacts to battery load, along with everything else!
I guess I'm lucky - I live and work in areas with generally good coverage and that offer reliable use, so I can comfortably treat this as a curiosity. I do realize however, that for some this is a major frustration with the device, and in that regard I hope we have some competent explanations and a solution as soon as possible.