8

So, often battery saving apps switch off WiFi when not connected to "save battery" (obviously). However, my question is this: if I was only out of the house for a couple of minutes, out of range of the WiFi, would turning it on and off be beneficial? In fluorescent light tubing, for example, leaving the light on is often more energy efficient than turning it off, only to turn it on again when returning a minute later. So, does the same apply to WiFi, or is WiFi being on less power consuming than turning it on and off? This might seem a stupid question, but – of course – I do what I can to save my battery life and in asking this question I am trying to better my battery saving techniques; there is nothing worse than your phone dying when waiting on a reply from someone. Is there any way to statistically prove either side or is it too "obvious" for analysis to be required?

Note: My phone is an LG Nexus 5 running Android 4.4.4.

EDIT: I believe the WiFi chip that the Nexus 5 uses is the "BCM4339 Wi-Fi Chip" manufactured by Broadcom, if this is of any help. Also, I am not looking for suggestions of other ways to extend my battery life, just an answer to my specific question. I have considered buying a power bank and other such products, but I'd preferably not have to carry these around, and simply have my phone's battery last an entire day.

Thank you for any contribution.

  • Off the main topic... I believe leaving fluorescent lights on doesn't save energy directly, it just saves tube life, which may indirectly save energy due to the energy cost of replacement. Using a car might be more accurate, as it is one of the few types of devices that use more energy to restart than turning it off for a short time. – Marty Fried Aug 8 '14 at 16:51
5

As you already pointed out in your question, it's a matter of intervals. Of course, turning it off-and-on like a cars indicator light is more power consuming than simply leaving it on – and turning it off for 12 hours does really save juice. The matter is to find the right "interval".

Not quite up-to-date anymore, but still sufficient for the "raw estimate": About two years ago I've posted a table with some consumption data, using two devices as reference. Let's average the values a little and assume they've changed "for the better", so we can say:

  • WiFi in standby uses about 10 mW
  • WiFi downloading uses about 800 mW
  • WiFi uploading uses about 400 mW

WiFi searching for a nearby AP to connect to, including the entire process, must be somewhere in between (no values in my table or their sources), but I'd assume at least 200..400 mW here – as like with cell signal, it needs to fully power up to scan the area for available SSIDs. Let's further assume the connect process takes about 5..10s, just to have some numbers to juggle with:

  • 5s × 200 ms = 1.000 units (minimum power-up consumption)
  • 10s × 400 ms = 4.000 units (maximum power-up consumption)
  • 10s × 10 ms = 100 units (maximum standby consumption)

By those numbers (being just raw estimates, not scientific calculations!), results would be:

100..400 seconds are the minimum "disconnected time" for your described "toggle" to not consume more juice than staying in standby. Toggling would thus make sense only for breaks of about 10 minutes and up.

For a similar calculation, you might be interested in 2G versus 3G: Does it really save battery? :)


To avoid misunderstandings:

As Dan correctly pointed out in the comments, I've omitted a lot of details here. I was aware of that: the above is nothing but an idealized number-juggling, which should show that even at best circumstances permanent toggling is no good idea. Calculating "exact numbers" which are at the same time "absolute" is impossible, as too many factors play a role here:

  • noone expects WiFi idling at 10 mW with the connection lost. It will certainly "power up" on a search for available networks.

Plus the points Dan mentioned:

  • many apps that perform network operations in the background use a broadcast receiver to run when the network connection comes up. If you're actually connecting to a network each time you turn the Wi-Fi on, all these apps will run, drawing more power. On each "toggle on and connect", that is – while they wouldn't do so at all otherwise, or at least at very lower intervals
  • if the Wi-Fi was on all the time, the same apps might run more or fewer times to update or sync data. This very much depends on the apps installed/used
  • if you're waking the device up and/or turning the screen on just to turn on Wi-Fi and check for messages, that'll use more power than the Wi-Fi itself.

So please read my conclusion as "it makes no sense for time-frames smaller than 10 min". The longer the interval, the more likely it makes sense – and the shorter, the less.

  • 1
    Anytime, Ben: You're asking interesting and good questions, which very much motivates me answering them (not the first time we meet ;) – Izzy Aug 8 '14 at 15:59
  • 1
    There's a big factor you're neglecting: many apps that perform network operations in the background use a broadcast receiver to run when the network connection comes up. If you're actually connecting to a network each time you turn the Wi-Fi on, all these apps will run, drawing more power. And of course if the Wi-Fi was on all the time, the same apps might run more or fewer times to update or sync data. And of course if you're waking the device up and/or turning the screen on just to turn on Wi-Fi and check for messages, that'll use more power than the Wi-Fi itself. – Dan Hulme Aug 8 '14 at 16:47
  • @DanHulme yes, this is a good point indeed, this is why I asked the question as the figures must be tight! – Ben Porter Aug 8 '14 at 17:11
  • @DanHulme Valid points all – though I didn't "neglect" but willingly "ignored" them. Still, you're correct reminding me of those. To avoid misunderstandings, I've updated my answer accordingly. Thanks for the "broadcast event", my "listener" was active: data received :) – Izzy Aug 8 '14 at 17:50
0

Turning the wifi on and off is more energy efficient than leaving it on all the time. I cannot think of a way to statistically prove this since the power consumption is influenced by other factors. For better battery life I can recommend greenify, turn off all location services and disable the "always allow network scan" in the wifi settings.

  • Thank you for your answer, I hope someone out there can prove this statistically though, as perhaps turning it on and off is less effective than leaving it on for, say, a minute? I shall check it out, thank you for your contribution! – Ben Porter Aug 8 '14 at 12:53
  • @BenPorter There you go, I've tried setting up some "fake stats" for you. Should be pretty realistic, though. – Izzy Aug 8 '14 at 15:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.