Monday, July 19, 2010

Apple Apologizes Reluctantly for iPhone 4 Antenna Problems, Blames Media, No One Explains Problem Plainly

Steve Jobs and the Apple spin machine are at it again with their latest press conference on the iPhone 4's antenna problems: http://events.apple.com.edgesuite.net/100716iab73asc/event/index.html. Steve's presentation makes it clear both implicitly (with his body language and dramatic pauses) and explicitly (with what he says) that he thinks there's no real problem here. He expresses his view that the only reason all this hubbub occurred is because the media is blowing the problem out of proportion.

Apple's successful manipulation of the media has led to some of the best free advertising for any tech company in existence. Now, the combination of media uncertainty about how to properly explain a legitimate technical performance issue with the iPhone 4 and a lack of clear technical explanation from Apple has led to negative coverage. Boohoo. Welcome to real world. The media giveth and the media taketh away. They especially take away when customers report a real problem and the company responsible tries to claim everything but responsibility for it.

You can't have it both ways Apple. This performance issue can't be both the result of an Apple design choice to expose the antenna and its "weak spot" on the outside of the phone and not your fault. The extremely limited data Apple provides in its presentation show an increase in dropped calls from the iPhone 3GS to the iPhone 4. Given that the "revolutionary" antenna design was supposed to improve wireless performance, seeing poorer performance is clearly an issue. Again, you can't credibly present those data showing an increase in dropped calls and then claim there is no technical issue.

As someone who knows just enough to be dangerous about radio signals and electronics and is willing to speculate based on the coverage I've seen so far, I think the first failure is that no one - not Apple, popular media, the tech media, or any experts brought in to comment - has succinctly explained the problem and its likely causes. Anandtech did a great job of proving quantitatively, as best as you can with something as complex as RF devices and without the complex test equipment, that there is a problem in their article: iPhone 4 Thoroughly Reviewed. Where they didn't do so well is in explaining it plainly and this poor explanation was repeated in follow-up interviews such as CNet's Reporter's Roundtable and TWIT (This Week in Tech) episodes. I can give the journalists, podcasters, and bloggers some leeway because they're not all experienced engineers and this poor performance - and its inconsistency among users - is likely due to a complex combination of technical issues.

As I've said before, its easy to criticize and harder to do better. Here's my attempt at doing better.
Caveats: Modern radios and the antennas that work with them are extremely complex. I understand antenna design is hard and explaining it is even harder. Its not possible to explain plainly something so complex without making some gross generalizations.

Background: Every cell phone manufacturer takes as many of the complex factors determining connection performance as they can into account in their phone design. Its a fact of life in the mobile wireless industry that you will have something nasty like a hand (in terms of causing RF signal attenuation) covering your antenna and blocking its ideal signal path to the tower. I'll call this Problem A. Manufacturers have handled this problem by making ever more sensitive receivers, better amplifiers, and moving antennas around inside their phones. Since modern handsets are about the size of a hand, the insides of the case provide limited options for improving performance by moving the antenna. The wireless standards bodies have addressed these realities by including all sorts of digital encoding and processing schemes into the communications spec such that weaker and weaker signals still result in a successful connection.

The simple part: Apple thought they could break the mold and do something revolutionary by putting their antennas around the outside of their phone, thereby reducing the effect of Problem A. Instead, they added a second problem, Problem B, that occurs only when the phone is held a certain way. When a user physically touches the two antennas running around the outside and electrically connects them, they drastically change the RF properties of the antenna-hand system and make the antenna less sensitive to the frequencies it uses to connect. Problem B reduces signal at the receiver which compounds with Problem A to make the signal drop even more sharply. Problem B simply cannot occur when the antenna is insulated inside the handset's case.

To put some numbers on these problems, Anandtech measured their compounded drop (A&B) to be 24 dB, or about 250x less signal than is otherwise received. They further measured the attenuation with the phone in a case, essentially isolating Problem A, to be 7.2 db, or about 5x less signal than is otherwise received. So, Problem B knocks the signal down roughly (250/5=) 50 times! Whereas internal antenna designs have to deal with Problem A, Apple's iPhone 4 antenna design has to fight the compounded result of both Problems A and B.

The complex part: So, why don't all customers see the same increased call-drop or poor connection quality behavior? Three reasons:
1) Not all users expreience problem B. They either don't hold the phone that way or they have a case which prevents it. They only experience Problem A. Therefore, they don't experience the extra ~50x drop in signal strength.

2) Digital encoding technology hides the issue in strong signal areas. As previously mentioned, and explained in Anandtech's article, a modern digital handset can maintain its connection at the same level of quality in the presence of an extremely low signal that is still above threshold as in the presence of a high signal. A signal weakened 250 times can still be above threshold. With digital technology, as long as your signal is "good enough," the connection stays good and the user notices no ill effects. When the signal level drops ever so slightly below the systems' threshold, however, the connection just drops. There's very little in-between where the connection sounds fuzzy or distorted like with older analog technologies. This technology effectively hides from users whether they're in good or bad signal areas for all situations above threshold.

3) Wireless phone users have become accustomed to calls dropping and intermittent connections. There are enough other problems wireless handset manufacturers and carriers face, not described here, to take up the rest of the alphabet and then some. AT&T has gotten its coverage and service heavily criticized of late, and I think there is some merit in that criticism. Regardless of AT&T's specific issues, all carriers face these problems, all drop calls, all drop connections. If you hear that AT&T's service is poor in your area and buy an iPhone anyway, then dropped calls and intermittent data connections are something you're expecting. What are a few more of something that has not previously been quantified?

No comments: