antenna-theory.com

[wirelessrudy]

How is the S/N ratio effected if we narrow the channel bandwidth for a radio link by half.
If hardware stays the same but we only narrow the channel bandwidth by half we see some increase in signal at the receiving point but what happens actually with the absolute noise level?

My point is that the S/N ratio improves due the stronger signal of the wanted radio wave at the receiver only. Others argue that also the noise picked up is less in absolute terms so the improvement of the SNR has two sources.

I am not ready to accept this since imho SNR is not changing as a result of having the receiver listening to a smaller amplitude of the carrier wave.
SNR is the ratio between the signal level of desired radio wave compared to the surrounding noise. If the signal level of desired radio wave would stay the same, the SNR would stay the same because I can't understand why a receiver listening to a radio wave with smaller amplitude is also picking up less noise than when he would listen to the wider band channel

I am referring to 802.11a/b/g/n technology with the use of standard atheros chipsets. It is known they don't have the best filters in the world when it comes to suppressing noise or cutting off the working frequency of use.

Can someone come up with some good and well explained argument(s) why I am right or wrong?

[bigSteve]

Noise does follow bandwidth.

That is, if you have a system with a 1 MHz bandwidth with noise power N_0, and then you decrease the bandwidth to 0.5 MHz, the noise will be roughly N_1 = N_0/2.

Noise is typically random and internally generated (thermal, for instance). This noise can be plotted as a function of frequency - it is fairly constant. So you look at the "Power Spectral Density" of the Noise, you see a flat line versus frequency.

To find the power in the noise, you need to integrate across the bandwidth. So as you integrate a constant function (the spectral density), you get the noise power - shorter integration yields less power.