Why would you NOT want to use interference robustness?

Discussion in 'Mac Basics and Help' started by motulist, Aug 4, 2006.

  1. motulist macrumors 68040

    motulist

    Joined:
    Dec 2, 2003
    #1
    When would you NOT want to use interference robustness? I never understood why it's even an option you can turn off, is there some reason to not always to have it on? Does it use more battery or processing resources or something?
     
  2. spicyapple macrumors 68000

    spicyapple

    Joined:
    Jul 20, 2006
    #2
    It reduces wi-fi range and strength, I think. You only need it on when someone's using the microwave, which isn't continuous. Leave it off by default.
     
  3. mad jew Moderator emeritus

    mad jew

    Joined:
    Apr 3, 2004
    Location:
    Adelaide, Australia
    #3
    Yeah, it sacrifices outright throughput speed for a more reliable connection. I don't have it enabled because my Macs don't struggle to connect to the network. :)
     
  4. motulist thread starter macrumors 68040

    motulist

    Joined:
    Dec 2, 2003
    #4
    Reduced range does seem to be the communal tribal wisdom everywhere I look, but I looked all over and could not find a single shred of documentation stating that that's the case. I assume that it's true based on the volume of people saying it, but everyone assumed that a bowling ball would fall faster than a marble (in a vacuum), but that turned out to be false.

    Wait, it reduces throughput speed? Everyone else seems to say it only reduces range. It'd be nice to hear an official word from Apple to clear up this confusion.
     
  5. mad jew Moderator emeritus

    mad jew

    Joined:
    Apr 3, 2004
    Location:
    Adelaide, Australia
    #5
    Throughput speed and range are related. For simplicity's sake, I guess the best way to describe it is in terms of the compromise of broadcast power.

    Think of the broadcast power as being the combination of throughput speed and total range. If we reduce the speed (go back to 802.11B only), then the range is extended.

    Now, if we add a third factor (interference robustness) then the values of the remaining two (speed and range) must diminish because part of the router's broadcast power is being used for the robustness factor.

    In other words, if you reduce range, then you're effectively reducing the throughput speed of the machine at that particular point in space, with respect to the router.

    Sorry if that just complicated things, but it's the best way I can think of explaining it. :eek:
     
  6. motulist thread starter macrumors 68040

    motulist

    Joined:
    Dec 2, 2003
    #6
    I see now, it's a calculus derivative thing. Throughput speed is a function of signal strength, and range really means signal strength per distance. I was under the impression that you either had a wireless connection or you didn't, and range meant the point at which you could no longer make that connection. But you're saying that range really means a diminishing curve of signal strength as distance increases and signal strength ~ speed. See, calculus is good for something after all! :D
     
  7. mad jew Moderator emeritus

    mad jew

    Joined:
    Apr 3, 2004
    Location:
    Adelaide, Australia
    #7
    Yeah, that's exactly right. :)

    The claimed (and theoretical) speed of an 802.11G router is 54Mbps at the router however this speed decreases as you get further away.
     
  8. spicyapple macrumors 68000

    spicyapple

    Joined:
    Jul 20, 2006
    #8
    Basically what mad jew is saying (in fewer words than necessary) is that interference robustness is really a bust. :p
     
  9. mad jew Moderator emeritus

    mad jew

    Joined:
    Apr 3, 2004
    Location:
    Adelaide, Australia
    #9
    Haha. :D

    Well, it's good if you use the computer in an area with very high interference such as an apartment block. It should mean less small-scale drop outs and network slowdowns. It'll give a more consistent connection which is great for IM, among other things. :)
     
  10. motulist thread starter macrumors 68040

    motulist

    Joined:
    Dec 2, 2003
    #10
    Apparently you explained it well, because now I understand. And seriously, when I was learning calculus I was not convinced that calculus thinking would have actual applications in the real world, but after learning it I see it popping up again and again. I guess all this education is worthwhile after all. ;)
     
  11. mad jew Moderator emeritus

    mad jew

    Joined:
    Apr 3, 2004
    Location:
    Adelaide, Australia
    #11

    Learning just makes me realise how much I don't know. :D
     
  12. motulist thread starter macrumors 68040

    motulist

    Joined:
    Dec 2, 2003
    #12
    I'm totally with you man. Right now I'm learning organic chemistry and other advanced science classes like that, and while I'm pretty sure that this is knowledge I will never need, it sure does give you a glimpse into just how deep the rabbit hole goes.
     
  13. d wade macrumors 65816

    d wade

    Joined:
    Jun 27, 2006
    Location:
    Boca Raton, FL
    #13
    sorry if this sounds dumb... but what exactly is interference robustness? and what does it do?
     
  14. gman71882 macrumors 6502

    gman71882

    Joined:
    Jan 12, 2005
    Location:
    Houston, Tx
    #14
    Did you READ THE POSTS ABOVE!!! :eek:
    :D
     
  15. EMKoper macrumors regular

    EMKoper

    Joined:
    Dec 18, 2002
    Location:
    Stafford, VA
    #15
    I think the data rate is governed by the hardware based on the signal strength versus the apparent noise level to maintain a small bit error rate (signal-to-noise a.k.a S/N in radio geek talk). If you increase the distance between the devices the signal strength decreases but the noise stays constant so the S/N ratio decreases unfavorably... in order to keep bit errors small, the data rate must be decreased (or you can boost transmitted power). An example of real hardware tests is attached taken from (link to a .pdf) Link [it is not clear what his x-axis is... looks like a relative loss term, but moving to the right means lower S/N]

    So maybe it isn't exactly a calculus thing... but more like S/N statistics and radio frequency coding schemes to maintain constant bit-error rates ...
    RangeSpeed.jpg
     
  16. iMeowbot macrumors G3

    iMeowbot

    Joined:
    Aug 30, 2003
    #16
    The robustness feature is explained in this PDF. It simply raises the minimum data rate so that packets don't take so long to transmit that they are sure to be hit.

    The drawback of enabling this is that you lose that lower data rate and won't be able to reliably to connect to a weaker/more distant signal (as opposed to a strong one with intermittent interference).
     
  17. PaulinMaryland macrumors regular

    PaulinMaryland

    Joined:
    May 17, 2006
    Location:
    Maryland, USA
    #17
    Range vs. Speed in BlueTooth

    In BlueTooth, the range. vs. speed inverse relationship is not so simple. You can specify them a-la-carte.

    BlueTooth devices are specified by Version and Class.

    Version denotes speed:

    Version 1.1 = 768 kbps
    Version 2.0 = 3.0 Mbps

    Class denotes range:

    Class 1 = 200 feet
    Class 2 = 66 feet (or something like that)

    Don't you hate how the larger number is better for Version, but worse for Class? Why do they do this?

    Anyway, you can buy a BlueTooth adapter that's Version 1.1 (slower) but Class 1 (able to exchange data farther).

    The most desirable BlueTooth adapters are Version 2, Class 1.

    Many, many listings on eBay and elsewhere screw up the description. The title will say "BlueTooth Version 2". But dig down, and you find that they meant Version 1.1, Class 2... or Version 1.1., USB 2.0.

    Unfortunately, at this time no one makes a BlueTooth printer adapter that's Version 2. By printer adapter, I mean one that uses the stout, Type B USB connector. A few printers, including some HP inkjets and my Epson PictureMate Deluxe, include a second USB port that's Type A (the long, flat connector found on your Mac or PC). In most cases, then, what this means is that even though your MacBook or MacBook Pro comes with BlueTooth 2.0, you can print at only 25 percent of that speed.
     
  18. Eniregnat macrumors 68000

    Eniregnat

    Joined:
    Jan 22, 2003
    Location:
    In your head.
    #18
    It's a little more than this, but iMeowbot hit the nail on the head!

    Interference Robustness is a protocol to dynamically break up packet sizes and for resending lost/damaged packet. It also makes the system more resistant to accepting a lost signal (gives the computer a longer time to wait before it accepts that a signal is lost, and "disconnects" from a wireless source). A lot of interference doesn’t come from just similarly cycled sources, i.e. a 2.3-2.5Ghz (I know that's quite a range, but I am a little tired to remember exactly what frequency 802.x.y specs are- I think 2.4Ghz,). Any way, harmonic distortions, i.e., regular cycling of a 60hz signal can interfere with regularity. Lengths of packets are dynamically changes the lengths of packets so that if a number of packets are returned as damaged or missing, it can request smaller less efficient chunks.

    I can add to this is that I have seen anywhere from a nominal performance hit to a significant one in regard to bandwidth. Along with dynamic packet sizing it also automatically adds levels of redundancy to information sent and received, as well as preforming error checking with increased accuracy.

    The system lengthens the time it takes before it reports a lost network signal, and reattempts to find the network and execute handshaking.
     
  19. EMKoper macrumors regular

    EMKoper

    Joined:
    Dec 18, 2002
    Location:
    Stafford, VA
    #19
    Sure, but the Blue Tooth devices in the categories you state each carry different waveform protocols and effective radiated power that enable shorter or longer range or nominal or higher BT transfer rate ... but the principle is still the same as with WiFi... need a S/N ratio of some level to keep the bit errors down to some level for a particular rate of data transfer. The various Blue Tooth "classes" are just representative of various radiated power options (Class 1 = 20 dBm, Class 2 = 4 dbm, Class 3 = 0 dBm). And the data rate increase at lower power levels are probably enabled by signal processing "gain" using a more sophisticated waveform coding ... BT v2.0 uses Phase Shift Keying modulation scheme in place of the Gaussian Frequency Shift Keying of basic BT v1.x.
     
  20. tuqqer macrumors member

    Joined:
    Jun 28, 2005
    Location:
    Boulder, CO
    #20
    I'm going to guess that what d_wade was asking is, "where do you find this Internet Robustness, because I haven't seen it while using my Mac." I'm pretty sure, gman, he'd read the posts.

    Internet Robustness is located in the pulldown menu when you click on the airport signal bars in the upper right hand corner of the screen.
     
  21. mad jew Moderator emeritus

    mad jew

    Joined:
    Apr 3, 2004
    Location:
    Adelaide, Australia
    #21
    Also, Interference Robustness is available to turn on or off (on by default) in the AirPort Admin Utility. Of course, you need it switched on in both locations for it to work. :)
     
  22. rafifreak00 macrumors regular

    rafifreak00

    Joined:
    Aug 7, 2006
    Location:
    Kansas. No, not on a farm.
    #22
    Wow, I didn't know you guys would get into graphs and factual data and all that crap over a small question like such... but boy was I wrong...

    heres my response...

    Keep it off.
     
  23. Chaszmyr macrumors 601

    Chaszmyr

    Joined:
    Aug 9, 2002
    #23
    Theory aside, I've experimented with interference robustness using strong, medium, and weak connections and never noticed a difference between robustness on and robustness off.
     
  24. mad jew Moderator emeritus

    mad jew

    Joined:
    Apr 3, 2004
    Location:
    Adelaide, Australia
    #24
    That's a good point. It won't have much affect on an internet user with an 802.11G network.
     

Share This Page