Posted by Cary Chin on May 2, 2011
This blog originally posted on the Low Power Engineering Community 4/14/11.
By now, you’ve heard about AT&T’s recently introduced feature on the iPhone 4 – along with iOS 4.3, the iPhone 4 tethering feature now allows you to create a “personal hotspot” that can be shared with up to 3 other devices. One interesting twist on the plan from AT&T: the additional $20 per month cost of the plan comes with an additional 2 GB of data, for a total of 4 GB per month. It seems as though all of the major carriers are pretty much settling on around $10 per GB at this point, and that seems like a fair price for now. Don’t look for new unlimited data plans anytime soon – the move to 4G (or at least 3½G) has already started, and carriers are looking for clean ways to get rid of all unlimited data plans – that makes sense, since the transmission speeds will soon allow you to use MUCH more than your “fair share” of wireless bandwidth on an unlimited plan.
I signed up for the personal hotspot option when it was introduced, finally replacing a jailbroken iPhone 3GS running MiWi as my portable hotspot solution. I still haven’t convinced my nephew that this was a smart or even reasonable move. In fact, I was very happy with the performance of MiWi, although I was never totally comfortable with the idea of jailbreaking my phone, both from the standpoint of getting new software updates and keeping the software stable, as well as violating the terms of agreement with AT&T. Now I’m happy to pay my $20 per month to have it all, and sleep soundly at night (actually that’s never been a problem).
But I digress…the really fun part about having the personal hotspot on my phone is that I can now run some additional power efficiency tests, and watch Star Trek a few more times! Recall that our previous testing running the 2 hour, 6 minute movie used about 1.1 Wh of energy on the iPhone 4, with the screen dimmed completely. Adding the screen and sound (which was negligible) pushed the total energy consumption to 1.7 Wh. And streaming the movie via WiFi bumped the total energy to 2.3 Wh. Our 3G streaming tests were even more interesting, resulting in energy consumption of 2.8 Wh to 5.3 Wh, depending on the strength of the 3G signal as measured by the number of bars displayed on the phone. I‘ve gotten a few comments regarding inaccuracy and instability of the numbers of bars method of measuring signal strength, but I’ve actually purposefully used these instead of the dBm numbers available in field test mode because I don’t want to imply any more accuracy than there is in the measurement. I’ve actually found that the number of bars is reasonably consistent for a general swag, and if anything, the resolution should only be viewed in three “quanta” rather than five. So I’m using 1-bar, 3-bars-ish, and 5-bars as my signal strength measurement. Anything finer implies more accuracy than I can measure consistently.
For my personal hotspot test, I’m streaming Star Trek through my iPhone 4 using the personal hotspot feature to connect with my iPad 2 via WiFi. This test was done in a different location, so I first ran a test on the iPhone 4 to calibrate everything. Streaming Star Trek via 3G in my room in Orlando (on vacation) with about 3-bars-ish of coverage with full brightness and sound took about 3.15 Wh of energy, the same as when I did the same test in my house in Palo Alto! The picture started out somewhat pixilated, but the 3G connection buffer caught up after a few minutes, and the movie generally looked great (note that my previous testing on the Verizon iPhone showed that under the same conditions the pixelization continued throughout the movie, suggesting that the connection bandwidth wasn’t high enough, and it was having trouble keeping up).
Next, streaming to the iPad 2 using the personal hotspot on the iPhone 4 – and it worked as advertised! I was able to watch the whole movie (2:06) with no problems. Again, the picture was a little pixelated in the beginning, and degraded at one point during the test, but in general the quality on the iPad 2 was very impressive. Final energy usage: iPad 2 used 6.5 Wh of energy (brightness and sound at maximum), and the iPhone 4 consumed 1.94 Wh to act as the bridge between the 3G signal and WiFi. Previous tests showed the iPad 2 using 6.25 Wh to stream the movie using Wifi so this was a little higher, but within a reasonable margin of error. Also, from previous tests, the iPhone 4 used 3.15 Wh to stream the movie via 3G at 3-bars-ish, and 1.7 Wh to play the movie locally – the difference of 1.45 Wh would be a good initial guess at energy to receive the data via 3G. Throw in the added cost of broadcasting the WiFi signal, and the 1.94 Wh seems to be just about right.
Not a bad vacation, and “The Wizarding World of Harry Potter” was fun, too!