There’s been a lot of talk about how signal strength and its representation by bars on cellphones, especially due to the iPhone 4 having a significant problem with attenuation when held the “wrong” way. I’ve also seen a lot of people posting responses to articles asking things like, “Why can’t they just represent the strength as a percentage?” and “Why isn’t there a standard to representing signal strength?”
The iPhone 4 has a unique design that eliminates some attenuation that the antennas would face in a normal phone because the antennas are not surrounded by a casing. That means the iPhone 4 can not only receive signals at a higher level, but it can also transmit at a lower level. The problem is that the antennas are on the edge of the phone, allowing a user to directly touch them. Since our hands are capacitive, bridging the antennas can cause a problem. If the antennas were recessed, the phone would probably be fine.
Apple’s first answer was simply to hold the phone differently. After complaints about the issue gained momentum, Apple announced that they were “shocked” to discover their formula for calculating how many bars to display was incorrect. Those of you with memories will recall that Apple changed the way it calculated the bars in release 2.1 of iOS (called iPhone OS then), artificially inflating the number of bars users were seeing. Now they’re releasing a minor update to iOS that will give users a more accurate and honest portrayal of the signal strength; unfortunately for them, most people will just see it as “This update made me lose one of my bars!” An appropriate solution to the hardware problem is to cover the antennas with insulating material (or, at the very least, the problem corner). This can cost pennies per phone.
People seem generally confused about what the issue is and why it affects some people and not others. Anandtech did a good review of the issue, finding that holding the phone in a certain way could result in a signal drop of nearly 25dBm. Decibels are logarithmic, so a single bel represents a 10:1 ratio. If you take an artificial number such as 100 to represent your signal, a 10dB drop would be dropping it to 10. A 20dB drop brings it to 1, so a 25dB drop brings it to less than one percent of what it was. That’s significant attenuation that will definitely cause problems in any area with a mediocre signal.
The other interesting side of this issue is how everyone is focused on signal strength, which is not a good measurement of reception quality. It’s kind of like the emphasis of X app store has Y apps. If you need some drinking water, a single jug of water is much better than getting an ocean of oil (perhaps courtesy of BP). When I worked on satellite communications equipment in the Air Force, outgoing signals were measured in watts (quantity because the quality is assumed to be perfect, and you don’t want to overload a receiver) and incoming signals were measured in BER which means bit error rate, how many good bits per each bad bit, measured in scientific notation (so a shorthand of “-7″ would mean 10000000 good bits for each bad one). It’s a much more useful measurement than milliwatts or even dBm.
All this means: 1) the average consumer does not have an understanding of radio propagation, 2) there are a lot of ways to measure a signal, and 3) there isn’t a universal standard. Not only is there no standard for what a given number of bars should mean, it’s extremely hard to define. You can say a given amount of power, but that doesn’t take into account the sensitivity of the receiver. The iPhone 4 is capable of holding calls that most phones would drop in a hands-off scenario. The other issue is that your real signal is constantly changing as your phone checks with multiple towers and the signal is affected by nearly everything, so you can’t simply show the current signal quality/quantity. You have to average it out for a short period of time to avoid the constant jumping.
Really, we need a more useful standard. The current bars are the equivalent of a fuel gauge: The reading is relative to your vehicle but not to another vehicle. You might have a full tank (five bars) and someone else might have half of a tank (3 bars), but that doesn’t mean you actually have more fuel (signal). Nor does it mean that you can go further than the other person because you don’t know the fuel efficiency (receiver sensitivity). Even with a useful measurement like fuel efficiency, you still run into different conditions that affect it like highway or city driving (cell tower transmission quality, interference, etc.).
It’d be nice to at least show dBm and have the color change to red as it gets closer to the receiver’s minimum usable level. For now, unless a universal standard is created for all cell phones, you would do better to measure your signal based on calls dropped, call quality, and data rate than by comparing bars.
