artless processes

By now you've no doubt heard of the apparent "design flaw" plaguing the latest Apple iPhone 4 that results in significant signal degradation depending on the orientation and positioning of the device in the hand. There are countless YouTube videos demonstrating the signal strength decrease that occurs when the device is held in a common hand position (cupped such that the lower-left edge is in direct contact with the skin of the palm). The video below shows how holding the device in such a manner results in a dramatic decrease in apparent signal strength via the signal bars.



Needless to say, people are demanding an explanation from Apple and even filing class-action lawsuits against Apple and AT&T for general negligence and misrepresentation of a product. After a week of complaints, Apple posted a letter on their website on July 2 responding to the claims of a physical hardware defect causing the problems. Apple asserts that the problem arises because "the formula we use to calculate how many bars of signal strength to display is totally wrong." The letter basically explains that they have been inflating the actual signal strength such that it appears there is a greater signal than there actually is. This is a reasonable tactic given the inconstant coverage and signal quality problems of the AT&T network. AnandTech performed an extensive review of the iPhone 4, including a mapping of the iPhone 4 Bars to Signal Strength. They obtained the relationship seen in the image below where each bar is mapped to a specific signal strength (in decibels).



A significant problem is immediately evident in the way the bars are being mapped to the signal strength, which is consistent with the Apple letter. The "full bars" scenario is mapped to having a signal of at least -51dBm while "1 bar" is mapped to having a signal of at least -113dBm. According to the AnandTech review, -51dBm is the best signal you can possibly get while -113dBm is the worst signal you can have before disconnecting entirely. This is fine and all, but over half of all possible signal strengths between -51dBm to -91dBm are being reported as the "full bars" scenario. This leaves an entire four bars to map the much smaller -91dBm to -113dBm range. What this means is that the iPhone 4 has an incredibly small dynamic range for displaying signal strength, which leads to the apparent inflation in signal the Apple letter blamed for the problems. The solution proposed in the letter would indeed improve the quality of the display of the signal strength by better mapping the intensity to the bars. It is important to point out that AnandTech review did agree with Apple's claims that the iPhone 4 has the best reception of all the iPhone models.

But there is one more thing...Apple's response does not completely address the problem people are experiencing. It is possible that in the video above, the person is actually in a region of questionable signal strength (say only -90dBm), which would fall within the 5-bar range. It is common for signal strength to be attenuated when any antenna-containing device is held because of interference from contact on the antenna functionality. This could cause the signal to drop anywhere from 1-20dBm, which could place the signal strength in the 1-bar range, which is what is seen in the video. This raises two issues, first being the severity of the drop (up to 20dBm according to the AnandTech report) and second the actual effect of the Apple software "fix" on the problem.

Experts are attributing the signal problems to the external antenna design of the iPhone 4. The iPhone 4 consists of two antennas that form the outer edge of the device, with one for Bluetooth, WiFi, and GPS and the other for UMTS and GSM (see image below).



External antennas offer both advantages and disadvantages over their internal counterparts (more). Because they are not confined within the casing, which is often an insulator or some kind, they can achieve greater signal strength. This has been seen with the iPhone 4 vs. previous iPhone iterations. The problem is that external antennas must operate in less controlled conditions that an internal antenna and are therefore more prone to unintentional attenuation (loss of signal strength) and detuning (changing the functionality of the antenna by introducing external factors). In the iPhone 4, these problems are thought to arise because of the two separate antennas that have been specially designed to coexist and function as the outer edge of the phone. However, when the two antennas are bridged by a conductive material (sweaty palm, skin, etc.) they are no longer separate entities electronically. Each is then operating with external factors (the body and the other antenna) that fundamentally change the tuning and functionality of the component. The side-view of the iPhone 4 below shows the small gap between the two antennae on the left side of the device that can easily be bridged.



It is because of these fluctuations from the expected conditions that the massive attenuation (and in some instances nearly complete attenuation) occur. The video below shows how the 3G network connectivity completely drops when held in the "death grip" with the iPhone 4 resting in the hand such that the gap is bridged by the palm of the hand.



Apple must have realized the potential difficulties inherent with utilizing an external antenna, more or less two in such close proximity that bridging was possible. The problem with the iPhone 4 design is that it almost seems like this was completely overlooked. Maybe it was the iPhone 3Gs disguise that the prototypes were field tested in that masked the potential complication. Or maybe Apple doesn't have enough left-handed testers or people that hold the phone in "weird" ways to pick up on it. It's interesting that a professor from Denmark's Aalborg University predicted problems with the iPhone 4 two weeks before it was released. Granted he based his hypothesis solely on the presence of an external antenna, not the interference caused by connecting the two together. Whatever Apple may attempt to rebrand it as, this represents a very significant industrial design flaw. Regardless of the signal strength being reported incorrectly, it is undesirable that such a significant drop can occur simply because an external antenna is utilized over an internal one. At the very least Apple should have coated the metal in an insulator to minimize inadvertent electrical conduction through the body.

This brings us back to the proposed "fix" by Apple. Obviously a software fix is the first and by far the cheapest and simplest solution. A hardware fix would require a recall of over 1.5 million iPhone 4s, which would be an unimaginable setback. I agree with Apple that the signal strength should be remapped to a more sensible scale such that a greater dynamic range is made available to more accurately portray the current network conditions. I think the downside of this will be that the AT&T network's shortfalls will become more apparent. Presumably the range covered by each of the five bars would become more standardized, which will result in previously 5-bar signals potentially being shown as 3 or 4-bar signals. It seems reasonable that any attempts to bridge the antenna will result in no more than a single bar (maybe two) of signal loss rather than the four or five that are occurring now. Apple will be masking the hardware problem by minimizing the effect of the signal drop by increasing the dynamic range available for reporting the signal strength.

In the end I seriously doubt anything more than this software fix will be implemented because it should essentially eliminate the problem from the point of view of the user. The AT&T network will probably seem weaker in places that were previously 5-bars because of the re-mapping but probably won't affect performance. Overall, the best solution are to hold the phone in a way that doesn't bridge the gap or just get a case that covers the entire region. Honestly why would you be walking around with a phone that has glass on two sides without some kind of case? I think it's also important to once again point out that the AnandTech report demonstrated that the iPhone 4 has the best antenna out of all the iPhones and that any wireless device can be held in a particular manner to impede its wireless communications. That being said, since the AT&T network is so crappy for making calls (or at least it used to be), there's no real harm in getting the iPhone 4 for all its other features (FaceTime doesn't even support 3G anyways).

Update
Looks like Applecare is already confirming that the software update will not fix the antenna problem on the iPhone 4. They are suggesting two solutions:

1. Hold the iPhone such that your hand does not bridge the two antennas
2. Buy a case

Obviously the customer must be making the error because Apple products are never flawed right? You shouldn't be walking around with an iPhone 4 without a case anyways so go get one and that'll (hopefully) fix everything.