What do cell phone reception bars mean?

April 9, 2007 7:59 AM

What exactly do those cell phone reception bars represent?

I seem to have great cell phone reception (Cingular) everywhere but my home. Yet often while I'm getting terrible reception and dropped calls, the phone tells me it has three bars of reception. What's THAT all about? Is my provider telling my phone to intentionally inflate the reporting of my reception or is something about my home causing my problem even though I nominally have good reception?

posted by norm to technology (8 answers total)

They don't mean much of anything, it turns out.

I don't know what they're displaying for GSM, but probably what they're displaying is the signal strength. For CDMA (which is what I know about) that's what they display, but in CDMA the signal strength is highly deceptive because it doesn't inform you of what the noise floor is.

The technical term is "EC/I0" (pronounced "ee-see-over-eye-naught") and it refers to the amount of the signal which is usable. In CDMA you can have strong signal (4 bars) and lousy EC/I0 and not be able to carry a call, and you can have low signal (zero bars) and excellent EC/I0 and carry a call fine. But they can't display EC/I0 because it fluctuates wildly (it could go from zero to four bars and back to zero again in just a few seconds) and would terrify users, so they display the signal strength, which at least has the virtue of being stable, though it doesn't really mean much.

Even worse... there is no industry standard for what "one bar" or "two bars" means. None. Everyone just sort of sets some thresholds, and even from the same manufacturer it can change from phone model to phone model.

Extrapolating from my CDMA experience, I would guess that in GSM they're displaying the signal strength of the paging channel, with an uncalibrated display not driven by industry standards.
posted by Steven C. Den Beste at 8:44 AM on April 9, 2007 [7 favorites]

The biggest reason why the signal strength isn't very helpful for GSM is what's known as "multipath". The signal reaching you from the cell can arrive by different routes which are slightly different lengths, meaning that they're offset in time. If you've ever seen "ghosts" on a TV when using rabbit-ear antennas, you're seeing the same effect. The "ghost" is a multipath component.

In GSM (or any other cellular system using TDMA) multipath is a form of interference. Strong multipath components can render the primary signal useless. The "signal strength" tells you how strong the primary signal is, but doesn't tell you anything about how much multipath interference you're getting.

Multipath by its nature is highly variable from one location to another. In some places moving just 2 meters can radically change how much of it you receive, thus drastically altering the ability of the phone to carry a call, without changing the signal strength in any way.
posted by Steven C. Den Beste at 9:07 AM on April 9, 2007

Two questions:

1. What brand of phone do you use? My wife and I had the exact same situation - crappy reception but only at home - except her Nokia worked well, and my Sony Ericsson and Motorola phones didn't work worth a damn. If you don't have one, but can get or test a Nokia, it might make a difference. From all accounts they have the best reception of any GSM phone.

2. Have you asked Cingular about this dead spot? They are usually very good about sending techs out to check for and try to resolve these issues. We moved from T-Mobile to Cingular specifically because T-Mobile just kept saying "the area you are in should have no reception problems" while Cingular said "we'll have a tech out there tomorrow to check reception."
posted by caution live frogs at 1:48 PM on April 9, 2007

This is my understanding based on working for a telco mobile data system several years ago. Please note that I was a programmer, not an RF engineer, so this may not be accurate.

The GSM standard does not specify the meaning of the signal bars on your handset (correctly known as the "signal quality estimate"). Each manufacturer uses their own formula to work out how many bars you see. This varies not only between phone makers, but also between models, and between firmware versions of the same model.

In short, you can't compare phones using signal bars.

You *can* - to a limited extent - compare the signal strength in different locations using the same phone, but even that isn't reliable.

What I'm about to describe is the typical behaviour for the Motorola devices I'm familiar with. Let me repeat, this may be out of date, and isn't necessarily typical of any other devices, but having said that, it's probably not far off.

When your phone is idle, the signal quality estimate (SQE) is calculated using some basic radio receiver parameters; the raw signal strength received from the cell tower, and the signal to noise ratio (the SNR). The SNR is by far the most significant component; in typical situations the indicator bars are an imprecise and slow-to-update SNR meter. In conditions of very low SNR the raw signal strength becomes a more sigificant factor in working out the SQE.

Note that this is entirely based on how well the handset is receiving the cell tower signal, and tells you nothing at all about how well the tower will receive your handset's signal; this is an important distinction because the two transmitters are very imbalanced in power; the cell tower can probably cut through a high noise environment with a high SNR; but your handset, which might be putting out an 18dB weaker signal (or even less) probably can't.

When you are actually on a call, the phone doesn't have to guess the effective signal quality - it's in two-way communication, and is constantly exchanging bit error rate (BER) information with the tower. In other words, it knows how reliably it is receiving data from the tower, and it knows how reliably the tower is receiving data from it.

This information is used to adjust transmit power at both ends; but it's also combined and displayed to the user via the signal bars.

If you've ever seen the indicated signal strength change significantly when you make a call, this is why - the estimate based on received signal strength is replaced by a fairly accurate (though completely uncalibrated) measurement of actual error rate.
posted by thparkth at 5:33 AM on January 13, 2008 [1 favorite]

Steven, your answer made it to Reddit's front page!
posted by growabrain at 12:48 PM on January 13, 2008

This answer also made it onto Gizmodo's Front Page and Boing Boing
posted by crazyray at 12:57 PM on January 13, 2008

In GSM (or any other cellular system using TDMA) multipath is a form of interference. Strong multipath components can render the primary signal useless. The "signal strength" tells you how strong the primary signal is, but doesn't tell you anything about how much multipath interference you're getting.

I don't know the particular details, but my understanding is that one of the things that makes CDMA better than GSM is that it can actually use the multipath signals to make your call clearer. This is why CDMA works even better than GSM in places where there's a lot of local signal echo - e.g.: cities.
posted by Caviar at 6:16 PM on January 13, 2008

For a nice overview on how GSM works, I recommend this page. He talks about signal power and how the handset handles a noisy channel; I used it quite often in my wireless-comm class, and found it to be a nice resource.
posted by virga at 9:00 AM on January 14, 2008