1
\$\begingroup\$

Background of my question:
I was checking the signal strength of wifi access points on my laptop in parallel with the internal Wi-Fi card and a nano usb wifi adapter connected to the same laptop. I saw that the signal strength measured on the nano usb wifi adapter was lower by 15 to 20 dBm as compared to the internal WiFi card, although measured almost at the exact same location.

My explanation for the behavior:
I assume that this is mostly due to confined space on the nano USB wifi adaptor compared to the internal Wi-Fi card, mainly limiting the antenna size, etc.

My actual question:
Unfortunately the drivers for both cards did not provide any noise measurements. However, now I would be interested what to expect for the measured noise in that situation. Specifically, how the measured noise on the WiFi radio is related to things like the antenna size/layout.
Naively, I would put the received signals on the WiFi radio into three groups:

  1. The signal itself, which is received via radio waves
  2. Outside noise that is also received via radio waves (for the 2.4GHz band e.g. microwaves, baby monitors, Bluetooth, etc.)
  3. Internal noise produced by the electronics inside the WiFi radio (operation amplifiers, thermal resistor noise, etc. as outlined here)

I assume that noise from group (3) will be independent for both WiFi adapters, as they will use different components and different layouts. From what I have seen in my experience, I expect them to be typically below -90dBm.
Am I right to assume that group (2) noise will also show up with different levels on my two Wifi radios, since also the actual WiFi signal strength is different.

Finally my theoretical hypothesis:
Assuming there is a babymonitor (from group 2) producing heavy noise on a 2.4GHz WiFi channel which shows up as noise around -60dBm on my internal adapter, could I then expect that this will cause a noise level of around -75dBm to -80dBm on my nano USB WiFi adapter, if also all WiFi signals (group 1) measured on the two WiFi radios typically differ by about 15 to 20 dBm. Thereby I assume that the internal noise (Group 3) for both WiFi radios is below -90dBm and not a factor for the measured noise value, which would be dominated by noise from group (2).
Is my hypothesis right? In other words, can I assume that the signal to noise ratio for both adapters will be the comparable, if only dominating noise from group (2) is considered?

\$\endgroup\$
  • \$\begingroup\$ Are they all allocated the exact same channel? \$\endgroup\$ – Andy aka Mar 19 '18 at 9:49
  • \$\begingroup\$ Yes, they are. Actually, both were connected to the same Access Point at the time of measurement. \$\endgroup\$ – oh.dae.su Mar 19 '18 at 9:51
2
\$\begingroup\$

First I would just like to clarify few things that could influence this and I think you are misunderstanding a few concepts.

1) Receiver efficiency

a) Due to the antenna: Just because your two antennas are in exactly the same location, does not mean they will receive the same amount of power. Antennas are all directional - it can be proven that a truly omnidirectional antenna is impossible. What this means is that every antenna is more sensitive in certain directions than it is in others. You can't know how every antenna was "oriented" and this could make a big impact.

To test this, what you can do is repeat the measurements with just one of the devices, but every time you do, rotate the device a little - both horizontally and vertically. As a result, you should find that the received power goes up and down depending on the orientation. If the different devices were not oriented the same way you will get different results - even if they were exactly identical in every other way.

In addition, not all antennas are equally efficient. Some antennas do a better job of "absorbing" the electro-magnetic wave and turning it into an electrical signal. A small antenna in a dongle is likely to be less efficient since it has to work in a smaller space and on a PCB, versus the "dedicated" antennas often found in laptops. This could add a few more dB loss, resulting in a weaker received signal.

b) Due to the receiver itself: Do you know where the power of the receiver is measured? Is it right at the antenna? Or perhaps after a first filter to isolate the wifi channel? Depending on how much there is between the antenna and the point of measurement, you will find that you might get different results. The manufacturer might have compensated for this, but you don't know. This could (and likely is) another source of error: there might be losses between the antenna and the actual power measurement, causing one to read lower power levels than the other.

2) Signal strength is different depending on location

You say that you measured in the same place, but how much difference is there? Even a few tens of centimeters could have a large impact because of the complex behavior of reflections you get with WiFi indoors. This could result in there actually being a significant difference in signal strength, even though they are measured in very close proximity. This video pretty much discribes what I am trying to say here: High Res Wifi Signal Mapping by user CNLohr on youtube

3) Errors in the actual measurement

How accurate are your measurement devices actually? Making an accurate system to measure power at these frequencies is not easy. Your WiFi system won't measure to single-digit accuracy. Especially if it is a cheaper dongle. So it could be that your dongle is just poorly calibrated, and this introduces an additional error.

Now to get to the noise

You have assumed that the received power could be split up into three groups: The desired signal, interference, and noise from internal circuits.

However, did not take into account the external background noise. Even if you are in a area where there is nobody transmitting - nothing at all, you will receive noise. In fact, if I just take a coaxial cable which has been terminated, and I connect that up to the receiver - it will see noise. This is thermal noise, and is due the thermal radiation of everything around us. Unless you are in a busy area, most of your noise will be of this type. This is always there, and you can't get around it. This is why the antennas of NASA telescopes use amplifiers cooled to very low temperatures, sometimes just a few tens of Kelvin.

For a WiFi channel (20 MHz bandwidth) at room temperature I believe this noise is around -110 dBm or so.

You ask

(...) can I assume that the signal to noise ratio for both adapters will be the comparable, if only dominating noise from group (2) is considered?

The answer is that nobody can tell you. Because of all the issues above, you can't know for sure what caused you to measure different results. If the interference noise is hitting one antenna in a direction where it is sensitive, it will see this signal as being "more powerfull" than the second antenna which is less sensitive in that direction. In addition, this baby monitor will not be constantly transmitting - the -60 dBm value you are getting is a bit meaningless by itself, without more information about how these devices measure them (is this peak received noise power? Is this average received noise power?). Hence, it is a bit difficult to talk about the "SNR" in this way. A final factor is the way they are all transmitting. If the baby monitor is connected to the same network, the WiFi protocol will actually try to avoid them talking at the same time. This illustrates the earlier issue of not knowing exactly how the power is measured and that "-60 dBm" figure is created - since each device will try and avoid transmitting at the same time as another device, they will not be a real noise signal you can treat like that.

\$\endgroup\$
  • \$\begingroup\$ Thank you very much for your answer. The Baby Monitor was just a theoretical source of noise, to better phrase my hypothesis. I was basically assuming to have an outside source of noise that is constantly transmitting in the frequency range of interest, but not using the IEEE802.11 protocoll. Regarding 1a), and 2) I will perform a few measurements this evening in order to rule out that the precise Position and Orientation have a big influence on the Signal strength. \$\endgroup\$ – oh.dae.su Mar 19 '18 at 10:22
  • \$\begingroup\$ If you were to have a real radiating noise source like that, you would have to factor in how each antenna is directive. But if they are identical in directivity then yes, the received noise power level would apear the same - or if one always measures 15 dB low, it will also read 15 dB low on the noise signal. \$\endgroup\$ – Joren Vaes Mar 19 '18 at 10:26
  • \$\begingroup\$ I also added a small bit on antenna efficiency, which I was going to do originally but forgot. \$\endgroup\$ – Joren Vaes Mar 19 '18 at 10:28
  • \$\begingroup\$ Hm, thermal noise happens inside the receiver. It's not externla. \$\endgroup\$ – Marcus Müller Mar 19 '18 at 11:58
  • \$\begingroup\$ @MarcusMüller everything made out of matter radiates thermal noise. If you point a noiseless antenna with a noiseless LNA connected to it to a wall at room temperature, you will still find you receive thermal noise. \$\endgroup\$ – Joren Vaes Mar 19 '18 at 12:01

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.