Slashdot Mirror

← Back to Stories (view on slashdot.org)

"Subconscious Mode" Could Boost Phone Battery Life

Posted by samzenpus on from the making-it-last dept.

cylonlover writes "University of Michigan researchers have proposed a new power management system for smartphones that could dramatically improve battery life. The system, known as E-MiLi, or Energy-Minimizing Idle Listening, addresses the energy waste that occurs when 'sleeping' phones are looking for incoming messages and clear communication channels. E-MiLi slows down the clock of a phone's WiFi card by up to 1/16 its normal frequency in order to save power, but then kicks it back up to full speed when information is coming in. The phone uses the header of the incoming message to wake itself up from its 'subconscious mode,' so the clock is at full speed to receive the main message. For users on the busiest networks, it could extend battery life by up to 54 percent."

85 comments

  1. fractions by Anonymous Coward · · Score: 4, Insightful

    by 1/16th? or up to 1/16th of? - - quite a difference there!

    1. Re:fractions by Anonymous Coward · · Score: 1

      Not just fractions.
      It doesn't "slow down by up to". My guess is it slows down up to 1/16 of normal clock rate.

      The way they wrote it makes it sound like the clock slows down no more than 1/16 of normal frequency. 1/16 clock reduction doesn't sound like it would yield "up to 54%" more battery.

    2. Re:fractions by ShanghaiBill · · Score: 1

      1/16 clock reduction doesn't sound like it would yield "up to 54%" more battery.

      Why not? "Up to 54%" just means less than 54%. So if it provided no improvement whatsoever, that would still be "up to 54%".

      I have up to a billion dollars in my pocket.

  2. Old ideas live again by dtmos · · Score: 5, Informative

    Patent 4,893,271, issued in 1990 and expired, covers an implementation of this idea in which the slow clock is a crystal oscillator, and the high-speed clock is synthesized, using the crystal oscillator as the reference of a phase-locked loop. It was used in tens of millions of Motorola radio pagers for exactly the reason stated in the article -- lower power consumption in sleep mode, while retaining the ability to process fast once a signal appears.

    1. Re:Old ideas live again by Hazel+Bergeron · · Score: 2

      And even in 1990 the idea that one could slow down the clock of a synchronous digital processor was obvious.

    2. Re:Old ideas live again by dtmos · · Score: 4, Informative

      Yes -- the trick is doing it without keeping the high-speed oscillator running all the time. In these kinds of embedded applications, the oscillator can draw a significant amount of power. Implementations that have a high speed oscillator running all the time, but divide it down while in sleep mode, draw more power than the invention, which has a low speed oscillator running all the time, and only generates high frequencies when they are needed.

    3. Re:Old ideas live again by Anne+Thwacks · · Score: 3

      I am pretty sure I worked for AT&T on digital pagers that did this well before 1990. We used a 4-bit micro and a modem that was mostly software. The entire protocol was designed to allow the pagers to "sniff" in time slots based on the pager address. I went on to work on "selective call radios" (cellphones) for Motorola which did the same thing in about 1988.

      --
      Sent from my ASR33 using ASCII
    4. Re:Old ideas live again by Anonymous Coward · · Score: 1

      which is also painfully obvious to anyone in the field. microcontrollers have had multiple clock inputs and internal oscillator blocks with support for switching between them built in for decades. You can also switch the low speed osc off while the high speed one is running if you're obsessive.

    5. Re:Old ideas live again by V!NCENT · · Score: 3

      This idea still realy sucks balls due to crappy WiFi signals never recieving anything. What we need is an expecting system. If you're not texting a while, just check once per 1,5min and do nothing else in between. Let it have a tickless OS.

      Oh wait, we already have that and it is called Linux (tickless)

      Should be a realy extremely obvious thing...

      --
      Here be signatures
    6. Re:Old ideas live again by dtmos · · Score: 2

      You can also switch the low speed osc off while the high speed one is running if you're obsessive.

      Yes -- if you're willing to pay for two crystal oscillators. The advantage of that patent is that you get by with one crystal.

    7. Re:Old ideas live again by dtmos · · Score: 3, Informative

      I am pretty sure I worked for AT&T on digital pagers that did this well before 1990.

      I went on to work on "selective call radios" (cellphones) for Motorola which did the same thing in about 1988.

      Quite possible -- the priority date for the patent is 7 November, 1983.

    8. Re:Old ideas live again by Dunbal · · Score: 2

      What, and force the telcos to invest in the infrastructure to do more work storing the messages and to take the load from all those incoming requests? NEVAR!

      --
      Seven puppies were harmed during the making of this post.
    9. Re:Old ideas live again by V!NCENT · · Score: 3, Informative

      While funny, SMS is being send over GSM, in the same package that is being exchanged with cell towers to maintain GSM (/2G) connectivity, irregardless. That package has empty space, so SMS (that's why it's limited by an amount of characters; so it can fit in that package) can be send without extra load on the cell tower buffer.

      What we are talking about here is stuff like Watsapp and Ping and that shit, which is stored on a central server anyway, so when cellphone asks server "What's up, man?" it then gets send without having to fill up the buffer, which is once per 1,5 minutes.

      --
      Here be signatures
    10. Re:Old ideas live again by MrHanky · · Score: 2

      No, this is something completely different. This is an even slower mode, set to watch only the headers of incoming packets sort of like a Slashdot system, in which people read the first couple of words of an article, then wakes up to comment on said article at full speed to get first post, just with the tiny tweak that it actually reads the rest of the message before replying.

    11. Re:Old ideas live again by philgp · · Score: 2

      -15 karma points for using the (non-)word "irregardless".

    12. Re:Old ideas live again by realityimpaired · · Score: 1

      I take it you've never left your phone off for a couple of days while people were texting you, eh?

      They already get buffered at the cell company, and all get delivered at the same time when you turn your phone on.

    13. Re:Old ideas live again by spottedkangaroo · · Score: 0

      It's getting popular with educated people as an ironic word to use. It's also, most definitely a word, it's just an uneducated one. I expect this use was the ironic one, not the uneducated one.

      http://oxforddictionaries.com/definition/irregardless

      It may surprise you to learn that many of the words we think of as totally acceptable were once like irregardless. Turns out, a few decades of ironic use (or even incorrect use) make it into a new word. I expect decimate will mean devastate shortly, if it doesn't already. I also expect FDA to be an acronym, if it isn't already. Those are the two that really bug me personally, but you pretty much have to let it go. The English you're accustomed to is also in use all over the world. I'm starting to hear some Indian idioms creep into US English from time to time. They make little sense to me sometimes. "Please do the needful." I'm hopeful it's some time before we're using that here in Michigan. WTF is that?

      --
      Imagine if you weren't allowed to use roads because a bus company complained about your driving 3 times. --skunkpussy
    14. Re:Old ideas live again by mbwjr12 · · Score: 1

      Many microcontrollers have internal oscillators that run at 32kHz for this exact reason, no second crystal needed. Additionally, you selectively control which peripherals are attached to which clocks, further minimizing power consumption.

    15. Re:Old ideas live again by vlm · · Score: 3, Interesting

      which is also painfully obvious to anyone in the field. microcontrollers have had multiple clock inputs and internal oscillator blocks with support for switching between them built in for decades. You can also switch the low speed osc off while the high speed one is running if you're obsessive.

      Its also painfully obvious to the RF engineering crowd. "obviously" if you have a multiplier chain the high freq mults are going to draw more power than the low freq mults. mults are never known for efficiency so if you have a couple stages you'll find most of your power is dropped in them not the osc. Also you wanna keep the crystal running all the time so its stable, constant temp, etc. Heck keep the first mult stage or so running to keep the load on the osc stable, if you want. That way the calibration output is theoretically usable all the time, instead of being FM modulated as you turn the mults on and off.

      Anyway, when you're actively TX or RX, then, and only then, power up the higher freq multipliers in the RX or TX chain (assuming you have 2 separate chains, you just got an instant 50% power savings at the cost of some complicated power wiring). If you're leaving the xtal powered up 24x7 to enhance stability, then powering down the multipliers seems terribly obvious... so that's exactly what they did in ye olden days. When I was a kid I fooled around with some even then obsolete motorola VHF TX and RX strips that had this "obvious" power saving feature.

      I don't have a cite, but pick up VHF/UHF/microwave oriented ham radio manuals from about half a century ago, you'll probably be successful.

      Merely terminating the last mult stage in a microprocessor clock input instead of a RF mixer is not exactly insightful or patent worthy. Its the equivalent of patenting the concept of using AA batteries in a flashlight instead of their traditional use in kids toys.

      The patent is probably much more specific to get around the staggering quantity of prior art. Probably specs the exact layer 2 framing protocol complete with diagrams of the sync header and stuff.

      --
      "Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
    16. Re:Old ideas live again by JerkBoB · · Score: 2

      It may surprise you to learn that many of the words we think of as totally acceptable were once like irregardless.

      That doesn't make them any less stupid. Anyone with even a basic understanding of prefixes and suffixes should flag "ir-" + "regard" + "-less" as nonsensical in context. "not" + ("regard" + "without") == "with regard", or worthy of consideration. It's right up there with "I could care less" in the retarded language department.

      For the record, I twitch at most uses of the word "decimate" as well. It's rare that one would get to use it properly in most conversation, and rarer still that it would both be used properly and understood properly by the recipient. I don't use it, personally.

      --
      A host is a host from coast to coast...
      Unless it's down, or slow, or fails to POST!
    17. Re:Old ideas live again by V!NCENT · · Score: 1

      You do realize that due to the lack of an official institute for the English language, there is no standard dictionary and grammer definition, rendering irregardless a perfectly valid English word, because it happens to be used by English speaking people.

      Thou art not upset by changing languages, art thou?

      --
      Here be signatures
    18. Re:Old ideas live again by JerkBoB · · Score: 1

      You do realize that words have commonly-accepted meanings, right? Your argument makes as much sense as me going around claiming that up is down and black is white and insist that because there's no institute of official English language my use of those words is perfectly valid.

      "ir" == "not"
      "regard" == "consideration"
      "less" == "without"

      ir- + regard + -less == "not (without consideration)" == "worthy of consideration"

      Yes, one could argue that "irregardless" is a valid grammatical construction, but it's invariably used to mean the opposite of what the construction would imply (to anyone who gives a shit about language and effective communication of ideas).

      --
      A host is a host from coast to coast...
      Unless it's down, or slow, or fails to POST!
    19. Re:Old ideas live again by maztuhblastah · · Score: 1

      While funny, SMS is being send over GSM, in the same package that is being exchanged with cell towers to maintain GSM (/2G) connectivity, irregardless. That package has empty space, so SMS (that's why it's limited by an amount of characters; so it can fit in that package) can be send without extra load on the cell tower buffer.

      No.

      Allow me to quote a relevant passage from a similar discussion elsewhere:

      THIS IS WRONG. THIS IS WRONG. A HUNDRED MILLION PEOPLE SAY THIS AND THEY'RE ALL WRONG BECAUSE THEY DO NOT FUCKING UNDERSTAND ANYTHING ABOUT GSM.

      Yes, it goes on the control channel (the same one used for handset registration and call setup), because that's the channel you can use without requesting that another channel be set up. No, it is not "free". No, it doesn't go in the "slack space" of another message that would be sent anyway. It sends a completely new packet on that channel (plus a reply that the message was accepted, plus more if it's a multisegment message or if there are any problems with delivery). And again, this is the channel which there is only one of, which carries the messages that allow handsets to join and leave the cell, allow calls to be made and answered, and allow all other data connections to be made. It's a finite resource, and if it runs out, the whole system stops working in that location.

      --
      The real litigious bastards...
    20. Re:Old ideas live again by spottedkangaroo · · Score: 1

      I'm particularly fond of "completely decimated" as if that were possible.

      --
      Imagine if you weren't allowed to use roads because a bus company complained about your driving 3 times. --skunkpussy
    21. Re:Old ideas live again by Bengie · · Score: 1

      +5

      I try reading 2-3 times before I reply, but this feature still tends to bite me quite often.

    22. Re:Old ideas live again by V!NCENT · · Score: 1

      Well, strictly speaking; you are correct.

      However, since we're not writing articles here and are communication with other preople over the internet; correctness isn't realy required.

      Then again, I do not live in an English speaking country, so I'm not getting annoyed about common language rape on a daily basis...

      --
      Here be signatures
    23. Re:Old ideas live again by Anonymous Coward · · Score: 0

      No, strictly speaking; he's incorrect.

      He changed "not (without consideration)" to "worthy of consideration". He should have changed it to "with consideration"

      "not without consideration" == "not (not with) consideration" == "with consideration"

      Irregardless of it's generally accepted meaning, I have no idea where he got "worthy of" from his original 'equation'.

    24. Re:Old ideas live again by russotto · · Score: 2

      Its also painfully obvious to the RF engineering crowd.

      Oh yeah, if they're so smart why don't RF chipsets include this feature right on the die? What's that you say, they DO? This is is merely the application of an existing technique to a problem solidly within its domain? Don't care, patent approved.

    25. Re:Old ideas live again by Anonymous Coward · · Score: 0

      According to the link you gave, it's an old 19th century word that is considered to be incorrect English.

      In other words, it's a word that uneducated people use; since "educated" people write using proper English.

      So there's no such thing as an "educated" person using the word irregardless.

    26. Re:Old ideas live again by fbartho · · Score: 1

      I don't usually get annoyed about the rape of the language, I'm more annoyed about the willingness of people to make themselves difficult to understand. Using words like irregardless grates, because I have to parse an extra syllable, then figure out they probably meant regardless and discard the whole meaning sense that I was halfway done appending to the sentence!

      Grrr.

      --
      Gravity Sucks
    27. Re:Old ideas live again by ynp7 · · Score: 1

      Irregardless, educated people often employ improper English ironically.

    28. Re:Old ideas live again by DRBivens · · Score: 1

      Yes, you can use "irregardless" in place of "regardless". You can also use "desalinization" and "preventative" instead of "desalination" and "preventive".

      The problem is, you're trying to communicate--to make your message understood by as many readers as possible--and these less-than-preferred forms can cause your readers to pay more attention to the language than to the message.

      Not to mention the lack of credibility many readers infer when they perceive an author can't even use the language properly.

      Personally, I'd rather improve my chances of being understood.

      --
      You have the right to remain silent. If you don't, anything you say will be misquoted and used against you.
    29. Re:Old ideas live again by BitZtream · · Score: 1

      If you're not texting a while, just check once per 1,5min and do nothing else in between. Let it have a tickless OS.

      You do realize that its just ticking on a longer period, right? Its not tickless.

      --
      Persistent Volume manager for Kubernetes - https://github.com/dwimsey/openshift-pvmanager
    30. Re:Old ideas live again by dtmos · · Score: 1

      All true. However, while internal oscillators have many good uses, their frequency stability is terrible -- usually something in the range of 0.1%, or 1000 ppm -- so it is difficult to use them when accurate timing is needed. For the application for which this invention was first used, paging receivers, timing accuracy of 100 ppm or so was needed, so a crystal was necessary.

    31. Re:Old ideas live again by Romberg · · Score: 1

      Ain't that the truth!

    32. Re:Old ideas live again by Anonymous Coward · · Score: 0

      You used it's as a possessive. Ha.

    33. Re:Old ideas live again by lsatenstein · · Score: 1

      Gee, I thought that at resonance, very very little power was required to sustain vibration. What am I not understanding about power loss and oscillators, unless the oscillators are mistuned from a natural resonance..

      --
      Leslie Satenstein Montreal Quebec Canada
    34. Re:Old ideas live again by Anonymous Coward · · Score: 0

      hey d00dz, i spose j00 liek this t00 then? there ain't no official institute for English, i kan rite h0w i want as long as someone else d0z too!

    35. Re:Old ideas live again by jasoncrowley · · Score: 1

      ir- + regard + -less == "not (without consideration)" == "worthy of consideration" That's how my minds eye reads the word as well, a hard reversal on "however/nevertheless" in a floppy sentence structure. The gibbering phrases looped back and forth from subject to subject, not unlike a wounded perch on a deserted beach. Irregardless, I attempted to continue on ignoring the cadence towards any meaning.

    36. Re:Old ideas live again by Anonymous Coward · · Score: 0

      You do realize that due to the lack of an official institute for the English language, there is no standard dictionary and grammer definition, rendering irregardless a perfectly valid English word, because it happens to be used by English speaking people.

      Thou art not upset by changing languages, art thou?

      Shut up troll. Every English speaking person who uses "irregardless" is too stupid to even know what it means. Once again, shut up.

  3. WiFi card...? by Half-pint+HAL · · Score: 1

    Presumably they mean the actual telephone tranceiver, and not an optional add-on such as wifi that can quite easily be turned on and off by the user....

    --
    Got them moderator blues I blieve I walk out the do', With these mod-points I been gettin', I 'most never post no mo'
    1. Re:WiFi card...? by mike260 · · Score: 1

      TFA repeatedly mentions WiFi.
      IIRC, the iPhone only enables wifi at all when you're actively using the phone - when it's sat in your pocket, wifi is off. Which would put a bit of a dent in that 54% figure.

    2. Re:WiFi card...? by brunes69 · · Score: 1

      This was my thought exactly. But the article REPEATEDLY mentions WiFi, which leads me to believe this research, while maybe novel, is not going to help most smartphone users.

      Even if you leave your WiFi on 24x7, it uses a FRACTION of the power of your HSPA/4G coverage. Power use of the 3G radio is the issue, not the wifi radio.

      I can leave my phone on wifi all day and if left to idle I will still easily have > 75% battery after 8 hours. With 3G, this would be well below 50%, maybe below 40% if a lot of data was coming in.

    3. Re:WiFi card...? by realityimpaired · · Score: 1

      If you have a decent data plan on your phone, or if you have very low data use (using your phone for e-mail only, for example, will probably not break 50mbyte/month), then there isn't really any point in having wifi on at all. I'll turn it on to update apps, because it's generally faster, but for casual surfing or e-mailing, HSPA (7.2mbit) is fast enough, and uses a fraction of the energy anyway, as most smartphones will already do what TFA is talking about with cellular data: switching to 3G when they need the speed, but generally sticking to 2G for stuff like e-mail push, SMS, etc. 2G uses a fraction of the energy that 3G does, and if I'm not playing games on the phone, it's not uncommon for me to get 3 or 4 days life out of it before I need to charge it.

    4. Re:WiFi card...? by Arlet · · Score: 1

      Also, it only saves power when you're actively using WiFi. Existing WiFi products already support a power save mode where the receiver is turned off completely for most of the time. It uses a timer to turn itself on briefly to receive beacons, and then turns itself off when there's no traffic indicated. This only leaves the radio on for a few % of the time. Additional power reduction from this mode will be very small.

      The new invention is only useful when you're using the WiFi at such an activity level that the WiFi power save mode would affect throughput too much. For most users, that will not be a significant amount of time.

    5. Re:WiFi card...? by Jerry+Atrick · · Score: 1

      The problem is very few mobiles properly implement the existing power saving modes, there's little chance they'll retrofit this new mode.

      The arithmetic doesn't work out in its favour either. This is essentially a low latency power saving mode. But for 18 hours a day the phone will be sleeping and so should its WiFi. For the 8 active hours you're unlikely to be continuously hammering WiFi and it should be in an existing power saving mode perhaps 6-7 of those hours. The power saving is likely to be 5-10% at best.

      It's an idea that could give a small incremental saving but would probably end up displacing one of the existing deeper saving modes and end up just improving latency for the same power cost.

  4. Just like the owners by Mike+Mentalist · · Score: 3, Informative

    Some of the text messages I receive read as if the sender was only subconscious.

    hw r u 2day? batt rnning low lol

    --
    I put my books on Amazon, Smashwords, Demonoid, ISOHunt and Pirate Bay. Search for 'Michael Cargill'
    1. Re:Just like the owners by Hazel+Bergeron · · Score: 3, Funny

      hw r u 2day? batt rnning low lol

      Sender is a midget artificial intelligence and is distressed that her bath is leaking.

      There's a hole in her bucket, dear Eliza.

    2. Re:Just like the owners by cerberusss · · Score: 1

      Sender is a midget artificial intelligence and is distressed that her bath is leaking.

      Actually, Steve Jobs removed the AI feature from the iPhone, because it used too much battery juice.

      --
      8 of 13 people found this answer helpful. Did you?
  5. What, weren't they doing this already? by Lord+Lode · · Score: 2

    Seems like common sense...

    1. Re:What, weren't they doing this already? by Anonymous Coward · · Score: 0

      Sure, but because someone else thought of it first, the rest of have to wait 20 years before using it

    2. Re:What, weren't they doing this already? by Anonymous Coward · · Score: 0

      obviously, that's why it was patended...

    3. Re:What, weren't they doing this already? by Asic+Eng · · Score: 1

      Well, at least in the automotive field that's a standard technique. The article doesn't describe anything particularly new. That might be the problem of the article rather than the problem of the research, though.

      Actually to some extent clock-gating is on the way out already: newer manufacturing technologies mean you get increased leakage - i.e. even when you don't switch the gates you draw a high percentage of the current anyway. So instead (or more likely in addition) of controlling the clock you go one step further and power-off parts of the chip. Power-gating is not as easy to handle as clock-gating, however.

  6. Low Power Listening by dominious · · Score: 1

    or LPL has existed in TinyOS for Wireless Sensor Networks for at least 4 years now. Check the BMAC and XMAC protocols for a start.

    These protocols actually turn off the transmitter power and wake it up every t milliseconds to check for activity. When transmitter is turned off for 95% of a second, then you gain 95% battery life when idle :) Ofcourse, there might be other issues in cell networks... I wonder if this can make another paper actually!

  7. SetCPU in Andoid Market by GuyFawkes · · Score: 2

    Does exactly this thing...

    --
    http://slashdot.org/~GuyFawkes/journal
    1. Re:SetCPU in Andoid Market by Anonymous Coward · · Score: 0

      I thought it was already built into Android? At least most hacked firmwares seem to do it.

    2. Re:SetCPU in Andoid Market by Anonymous Coward · · Score: 0

      It appears that SetCPU only manipulates the CPU clock. This article talks about reducing the Wi-Fi clock.

    3. Re:SetCPU in Andoid Market by Anonymous Coward · · Score: 0

      Yes but not for long if the company carries setcpu to court for patent infringement. First to file wins.

    4. Re:SetCPU in Andoid Market by kangsterizer · · Score: 1

      CPU clock != Wifi clock

  8. One Better.... by PowerCyclist · · Score: 0

    There's also an 'off mode' that will greatly extend battery life. Since we all know we're not turning our smartphones off, except to reboot them, I'm glad this feature is being developed and am shocked it's not standard practice yet. Than you to the other /. readers who pointed out the apps for controlling these functions.

  9. AND THE MAN IN THE MOON !! by Anonymous Coward · · Score: 0

    Little boy blue !!
    And the silver spoon !!
    Cat's in the cradle !!

    Subconscious your bedtime stories on that for a while !! It could do something for you !! Then again, it could do nothing at all !! It could do something for some, and nothing for some !!

  10. Huh?? by Anonymous Coward · · Score: 0

    Isnt this how most devices saves power these days (And all other days the last 30 years)?

  11. Know history or repeat it, your choice! by Anonymous Coward · · Score: 0

    Duh, Since the invention of CMOS, a clock = N electrons.
    Less clocks, less electrons.
    Filter coefficients must change to reflect the new delta time.
    Go back to school.

  12. ow about a "not suck" mode by Lumpy · · Score: 1

    Honestly ALL The phone OS's suck at one thing. Multitasking.

    Why cant i white list what multitasks and everything else is forced to close down when I close it? there are very few apps that I want running all the time. No I do not need my shopping list and angry birds running in the background. I want them crashed closed.

    There are a total of 5 apps that I would want to ever stay resident and running in the background, everything else I wand dead.

    Why cant they do this? it would increase performance and remove the damned problem of some apps fighting each other (iOS camera apps will fight, Android camera and bluetooth apps will also fight)

    How about fixing the damn bugs that cause battery waste in the first place?

    --
    Do not look at laser with remaining good eye.
    1. Re:ow about a "not suck" mode by bill_mcgonigle · · Score: 1

      Why cant they do this?

      Real OS's as smartphones are only a few years old. Something like systemd on Android could implement something like this (that is, all the parts are available).

      But look at the haphazard way Google built Android - they're cleaning up now but time-to-market was the only priority when the started.

      This, along with the battery life issue mentioned in TFA and missing encryption, is one of the reasons I still carry a 4-day-battery 'feature' phone.

      --
      My God, it's Full of Source!
      OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
    2. Re:ow about a "not suck" mode by c6gunner · · Score: 1

      This, along with the battery life issue mentioned in TFA and missing encryption, is one of the reasons I still carry a 4-day-battery 'feature' phone.

      Android does full-disk encryption (at least 3.0 does - not sure about earlier versiona) and battery life is much improved starting with 3.1 thanks to better CPU throttling and implementation of WiFi power-saving features.

      As for the original complaint, there are various apps which can kill tasks either manually or by setting specific criteria. Yeah, it would be nice if the OS had better support for it built-in, and it porbably will eventually, but it's still not an insurmountable problem even now.

    3. Re:ow about a "not suck" mode by gmuslera · · Score: 1

      You should try the (2 year old by now) N900 for good multitasking, it is still better in hardware than a lot of newly announced android and w7 phones, and had been optimized and stabilized a lot with the fixes that came all this time.

      Would be nice if one of the next community updates includes this method to save battery as an OS patch.

    4. Re:ow about a "not suck" mode by Anonymous Coward · · Score: 0

      Ugh, as a new developer to Android.. its issue is with being too easy to close activities (the building blocks which make up apps)! They are paused as soon as you enter another activity, and are killed after a while of being in the background (you have to manually specify an object to be passed back to you when the activity is recreated to make it look like it's how the user left it on restart). It makes it very difficult to maintain state!
        Check out http://developer.android.com/images/activity_lifecycle.png

    5. Re:ow about a "not suck" mode by Anonymous Coward · · Score: 0

      Or the N9 which is coming out in the next few weeks (depending on country) which runs a significantly updated version of the N900's OS.

  13. from the horse's mouth by Gravis+Zero · · Score: 1

    original article: http://www.engin.umich.edu/newscenter/feature/smartphonebattery/
    research paper: http://kabru.eecs.umich.edu//papers/publications/2011/xyzhang_kgshin_mobicom11.pdf

    --
    Anons need not reply. Questions end with a question mark.
  14. New dogs learn old tricks by mschaffer · · Score: 1

    Well, I guess new dogs can be taught old tricks.
    Clocking down a processor to save power while waiting for something to happen is hardly anything new.
    I am just surprised that smatrphone developers have just learned how to apply this to what they are doing.

  15. Wifi? by Anonymous Coward · · Score: 0

    Most phones do not have that running most of the time anyway - my current favorite, XV6900 (AKA HTC Touch/Vogue) does not even have wifi. Hoping that meant cellular radio since that is what runs down most phone batteries.

  16. I have an idea by rossdee · · Score: 1

    its called the 'off' switch. Maybe I should patent it.

    1. Re:I have an idea by Anonymous Coward · · Score: 0

      So, this "off switch" of yours, does it still let you receive incoming wifi packets like this solution does?

  17. Obligatory BR by starglider29a · · Score: 1

    Would this allow Androids to Dream of Electric Sheep? (Too obvious, should have been in the tags.)

    1. Re:Obligatory BR by treeves · · Score: 1

      That's not BR, that's PKD. BR never mentions electric sheep.

      --
      ...the future crusty old bastards are already drinking the Kool-Aid.
  18. Re:Decimate by DocSavage64109 · · Score: 1

    As Decimate means "removal of a tenth", wouldn't the proper use be dependent upon the original force being a multiple of 10, and exactly that 1/10th being removed? It seems like this would be rather rare.

  19. 802.11 (Wifi) has a sleep mode already (FYI) by kangsterizer · · Score: 1

    Wifi clients can have 3 stats:
    - active (all normal)
    - sleep awake (aka powersave awake)
    - sleep asleep (aka powersave asleep)

    When the client is in power-save-asleep mode, the access point buffers all the data going to the wifi client.
    Meanwhile, the AP sends specific beacons to the client, called TIM.

    From time to time, the client goes into powersave-awake mode to get those TIM beacons. It needs a lower power than active mode for the beacons as bandwidth is not a requirement. If the TIM says the client needs to fetch some data, the client wakes up to active mode and receives the frames. Else the client goes back to powersave-asleep.

    Now what does it means in the real world: it means the wifi is nearly 100% powered off when in powersave-sleep (even less power consumed than in E-Mili state I presume since the wifi radio chip is completely turned off), and probably only consume a little more in powersave-awake mode (E-Mili seems to operate at an even lower rate than wifi's 1MBit minimum)

    I'd like to see E-Mili vs Powersave real world measure consumption, but I am worried the E-Mili author just discarded the use of Powersave in his measurements.

    1. Re:802.11 (Wifi) has a sleep mode already (FYI) by Anonymous Coward · · Score: 0

      Since the ADC+RF budget for receive mode is typically 50% of the total power consumption, saving 54% is an impressive claim.
      I think these guys are suggesting to subsample the preamble somehow, explaining why modifications to the standard are required. Subsampling the preamble would permit the ADC convertors to run slower, saving power.

      Even if it works, not sure how it's going to save substantial power unless and address field is included. My WIFI VOIP device sitting at home would be constantly waking up in presence of the AP.

      I think their work is centered around solving the incoming VoIP call, so some of the chatter about 1.5 minute poll periods for text messaging don't apply

  20. Wake on LAN? by HTH+NE1 · · Score: 1

    So this is Wake on LAN, but for WiFi? Wake on WiN? No, theirs was better. Though I'd drop the hyphen: "EMiLi" can more easily be pronounced Emily while "E-MiLi" forces the long-E, giving E-Millie or, worse, if HiFi/WiFi/MiFi carries the day, E-My-Lie (WiDi is already unfortunate: "Why Die").

    --
    Oh, say does that Star-Spangled Banner entwine / The myrtle of Venus with Bacchus's vine?
  21. Title reminds me of the old Little Johnny jokes. by mj1856 · · Score: 1

    Teacher: "Little Johnny, could you please tell the class about the subconcious mode on your phone? Be sure to tell how it affects the life of the battery." Little Johnny: "When I beat you into subconcious mode with my phone, I'll surely get life in prision for battery!"

  22. Prior Art by Fned · · Score: 1

    Energy-Minimizing Idle Listening... E-MiLi slows down the clock of a phone's WiFi card by up to 1/16 its normal frequency in order to save power, but then kicks it back up to full speed when information is coming in...

    That sounds pretty much like how I make it through Thursdays.

  23. Frequency scaling by Trogre · · Score: 1

    This looks to me like CPU frequency scaling, which has been common on desktop PCs since 2004.

    Having said that, reducing your CPU by a factor of 16 does seem much more impressive than the 2-3 which is usually employed.

    --
    "Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife
  24. Subconscious Emily by Anonymous Coward · · Score: 0

    'tis Dr Freud all over again.