What does that mean? How can you tell which signal its using? What attenuation?
Basically describe the test you did and how you came up with this statement. You haven't provided anything to backup the statement.
Well, first of all, it's not incumbent on me to "backup the statement." I was telling you my theory.
if you must know, I have a microcell. I have two 3GS's and a 3G that I've used with the microcell. None exhibit any weird behavior.
I received two iphone 4's on wednesday. Both switched from 5 bars microcell to 0 bars and no microcell within 20 seconds of touching the lower left-hand corner. This happens even 3 feet from the microcell. I had one iphone 4 replaced (it had other problems - no sound from earpiece). It exhibits the same problem with the microcell. You can tell whether you are using the microcell or not because the phone switches its status from "AT&T M-Cell" to "AT&T."
At my house, with the microcell off, I get 1 bar, typically, with the 3gs's. With the 4's I get 2 bars. If I touch the bottom corner of the 4's it drops to 1 bar, but keeps the connection.
With the microcell on, when I lose bars, it also drop calls.
It has been published that the prioritization method changed in OS 4.0 so that it favors reliable cells over strong cells. (Don't bother asking me to supply a link. Google is your friend).
I don't know what metric it uses to determine reliability (perhaps it keeps track. perhaps it has a database. perhaps the cell tells it - in some of those scenarios it's easy to imagine that it doesn't realize that microcells are reliable because someone forgot to tell the software that's the case), but if I choose to believe this change in prioritization, the behavior I see is consistent with the phone not realizing that microcells are reliable. If it sees a nice consistent connection on the microcell it has no reason to switch away. Touching the antenna causes a slight signal attenuation. This is consistent with behavior I see in other locations, such as at my office or on university avenue in palo alto, where touching the corner causes about 1 bar drop and about a 10-15% data speed reduction.
The attenuation is interpreted by iOS 4.0 as meaning "this may be a strong signal, but it smells unreliable to me. I'll keep dropping bars and comparing to other cells until I find a signal whose weighted strength + reliability is higher than the signal I have now.
The problem is that the weighting of the formula is all screwed up.
Anyway, that's my theory. And while I have a Ph.D. in electrical engineering, I have no expertise in antenna design or RF communications, so it's just a theory based on my experimentation with three iphone 4's, 2 iphone 3gs's, and one iphone 3g.