Some of my friends asked whether they should use Wi-Fi, Bluetooth or USB when using their iPhone’s cellular data for Internet access on their computers. In other words what is the best connection to use for Personal Hotspot. At a glance, Wi-Fi tethering should be faster than Bluetooth because of it’s higher theoretical data transfer rate. But using Wi-Fi for personal hotspot tend to drain your battery faster. Of course USB has the fastest transfer rate of the three but you need to carry the Apple’s 30-pin USB cable everywhere.
I’ve done some tests and concluded that you should prefer Bluetooth over Wi-Fi for personal hotspot. Why? Two reasons:
- There isn’t any significant drop in real data transfer speed when you use Bluetooth for 3G cellular data access. This is because the typical 3G data plans gives significantly lower real throughput than what Bluetooth normally offers. I’m not talking about theoretical transfers here – I’m talking about the the real kilobytes per second of download and upload to your favorite Internet sites.
- Bluetooth tethering can be initiated even when the iPhone is in standby with the screen turned off. Unlike Wi-Fi where you need to turn the phone on on when you need to tether. It’s a whole lot more convenient since you can keep the phone in your pocket and just initiate the connection from your laptop or iPad.
Of course it’s better to use USB tethering when your laptop is connected to an AC adapter since you can charge both the phone and your laptop at the same time. I carry a short (5 cm) iPhone-to-USB cable in my laptop bag just for this use.
I did some tests of the connection speed of M1‘s cellular data connection performance and compare it untethered on the iPhone, tethered to Mac and tethered to an iPad. All tests were done by the SpeedTest application, with SpeedTest.net for iOS for testing on iPhone and iPad and their Flash-based version for testing on the Mac. Each test was done four times on each device which amounts to twenty test cases in total. Test devices were iPhone 4 (the older one), iPad 2, and Macbook Air (2011). It was done on a Thursday night at around 20:00 local time.
Bluetooth tethering has slightly lower response times than Wi-Fi when tethered to the Mac and iPad. However the variance is high and maybe this really need a re-test with more samples (which I don’t have the inclination to do at this point). The graph below shows the average response times in milliseconds and the error bars represent one standard deviation above or below the mean. In other words, the range within the error bars represent 68% probability assuming that the results were normally distributed.
However data transfer tests doesn’t show a significant difference between Bluetooth and Wi-Fi tethering. In the graph below you can see the average upload and download speeds between the various devices. The error bars represent one standard deviation or 68% probability of more test results will fall into that range. All values are in Mega-bits per second (there are eight bits to a byte and 1024 Kilobits per Megabit).
You might be curious, what was the data plan that I subscribed to and used in these tests? I used the M1 iPhone Value plan that promises “Up to 14.4 Mbps” data transfer speed. You can see for yourself that M1′s real data transfer of about 0.15 Mbps is vastly below the speed that they advertised – which is by the way the theoretical maximum of 3G data (that anybody can conveniently quote without testing nor needing to commit to an SLA).
You can test these out yourself and see whether your cellular operator’s real data transfer rates are anywhere near Bluetooth 2.1′s theoretical limit of 2.1Mbps. Even if your favorite mobile operator is twice as fast as mine, you’ll come to the same conclusion that Bluetooth is the better choice for wireless tethering.