1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Noise Floor

Discussion in 'Tomato Firmware' started by adlerfra, Dec 14, 2010.

  1. adlerfra

    adlerfra LI Guru Member

    I clicked on the Measure Button on the Device List screen and my 3 wireless devices were:
    Before:
    RSSI Quality
    -53 46
    -28 71
    -56 43
    After:
    -54 28
    -29 53
    -55 27

    I was surprised there was such a large change in the numbers. Is this measurement accurate? The Noise Floor is: -82 dBm.
     
  2. maurer

    maurer LI Guru Member

    does <1dBm seems "such a large change in the numbers" ??
     
  3. adlerfra

    adlerfra LI Guru Member

    You are right about the RSSI. But why did the Quality drop so much?
     
  4. Toastman

    Toastman Super Moderator Staff Member Member

    Until you measured the "noise floor" the router worked on the basis of the default setting of 99 to calculate the signal to noise ratio (quality). [Quality=the difference between the noise floor and RSSI readings.]

    There was an increase of 17dB in noise floor. And your "quality" figure changed by around 17. It's all correct.

    The only thing that is curious, is that your noise floor is horrible. Since the really important thing in any radio link is the signal to noise ratio, then you might try to identify why your noise floor is so high and improve it. In your case, you don't need to go overboard on this because your "quality" figure is still quite high, and you should still be able to get a 54Mbps link - but other people with lower quality readings may need to do something.

    Are you running your routers on weird channels, and therefore identifying other routers on 1,6,11 as noise? Try changing channels, stick to the same ones that are in use around you, never try to go in between, thinking those channels might be better. The link below explains why.

    Is there anything else that could be causing interference to your routers. Microwaves, baby alarms, RF security cameras ?

    Use the latest inSSIDer to take a look at the overall picture. The picture it gives is worth a thousand words!

    This link may be useful

    http://www.linksysinfo.org/forums/showpost.php?p=367889&postcount=7
     
  5. Dudeman456

    Dudeman456 LI Guru Member

    Thanks Toastman for the information. I was wondering if you could shed some light on the following.

    [​IMG]

    The SSID of "The Future" begins before the channel set, and end at channel 7.

    Also a few times there were two towers side by side that reached the top of the chart.

    Noise Floor -91
     
  6. Dudeman456

    Dudeman456 LI Guru Member

    Just after I put up the first picture.

    [​IMG]

    My own wifi only reaches -30 on the chart, how are others able to be able produce more power than my own router 7 feet away?
     
  7. Toastman

    Toastman Super Moderator Staff Member Member

    That reading is nonsense, the maximum figure that inSSIDer can produce is 50. It's a bug.

    Now, about the plots. Well, it's very clear that

    1) The router on channel 3 can cause interference to routers on both channel 1 and 6

    2) Another one on channel 9 can interfere with channels 6 and 11

    3) The double-channel N router (the future?) will cause interference to nearby users on channel 6. But actually, it **should** detect routers on that channel and simply refuse to operate in double channel mode. That function doesn't seem to be working, does it? It may not be hearing those other routers, they may not be visible where it is located and therefore it could be OK, but we have no way of knowing that.

    All of these routers are very weak at -80, or worse, you shouldn't have too much problem with them. You say your own router reaches -30, but INSSIDER doesn't give output over -50 - refer to their website on this. This does limit INSSIDER's usefulness as a tool, but that's a limitation of the RSSI measurement system on wifi cards.

    Is your router on these charts at all? I guess not.

    Anyway, this is a good illustration of how little people understand how to use the channels, two routers on "wrong" channels in your own small area and one dual channel N router overlapping the majority of users on channel 6. Not good - but typical. You can see why people are getting angry about N wifi on 2.4GHz !
     
  8. Dudeman456

    Dudeman456 LI Guru Member

    My router(System) should be on the charts, but during the sampling, it was experiencing something I reported many times on this forum. Where it would just disappear.

    Before using inSSIDer, I had my router on Channel 1(Noise Floor -92), where it showed a slightly higher signal than where I have now on Channel 11(Noise Floor -85).

    [​IMG]

    From your analysis, would you agree that Channel 11 is my best choice under the conditions?
     
  9. Toastman

    Toastman Super Moderator Staff Member Member

    From the chart only, I would have picked channel 11. From your noise floor readings though, I would modify that. I'd choose the channel with the best noise floor out of 1,6, or 11. In other words, find the best signal to noise ratio. Channel 1 would seem better based on your posts. Try both and do a speed test from time to time with a file transfer. Ultimately, that's what matters, in fact it is the only thing that matters to you!

    Unfortunately, we do have another problem which as an RF engineer I am only too well aware of - but rather loathe to introduce as it will upset a lot of people's pet assumptions. And that is - that the receivers in wifi cards can't really measure signal strengths or signal to noise ratio in the normally accepted ways. They use software algorithms to come up with a guess. These algorithms work on data transmissions, but they probably don't see on-channel interference from non-data sources (baby alarms, video, microwaves) at all. These receivers are not communications receivers by any sense of the word - they are extremely simple devices enhanced by software. So we have a real hodgepodge of assumptions used in these cards, every manufacturer's hodgepodge is different, and it's actually a wonder they work as well as they do. The INSSIDER guys are engineers, and they also know this. You can see some of their thoughts being expressed here:

    http://www.metageek.net/forums/show...timate-Signal-to-Noise-Ratio-(SNR)-discussion

    Another point - if a router is too close to the client, quite often one or both receivers can overload and this results in higher error rates and slower throughput. During some of my tests with higher transmit powers, I found that some clients were OK at 2 metres away, others needed 3 or more. This may or may not have something to do with the "disappearing signal". Just a thought.

    I wonder what your wifi card is that you are using with INSSIDER and pondering, why you are getting an output above -50. Curious!
     
  10. Dudeman456

    Dudeman456 LI Guru Member

    The disappearing signal must be a localized router event, as affects all devices connected to it. Different wifi brands and all, there has only been one instance where all the devices disconnected but one stayed on.

    As for the wifi chipset, its an Atheros AR5008X.
     
  11. Dudeman456

    Dudeman456 LI Guru Member

    I did a quick scan with a Ralink card, and experience a reading of -39 for my router.


    [​IMG]
     
  12. Toastman

    Toastman Super Moderator Staff Member Member

    That's very interesting, because according to INSSIDFER site, it shouldn't, and it it never has before - there was a prior version that did show above -50 but that was a couple years ago. Maybe they fixed it, but my card is still stuck at -50. However, it does also occasionally peak to a random number, but it never stays there.
     

Share This Page