Some questions about the bandwidth noise

Hello everyone,
I know that the avaiable thermal noise power spectral density at the receiver input is approximatly -173 dBm/Hz (at room temperature).
I've always read in the literature that to calculate output signal and output noise I have to do as follow:

Signal_output = Signal_input + Receiver_gain
Noise_output = Noise_input + Receiver_gain + Receiver_noise_figure

Where I think "Noise_input" is the thermal noise power spectral density (-173 dBm/Hz).

I don't understand... Does not the signal input have its own noise??? I mean, the signal input is generated by a source, such as a RF transmitter. If the transmitter is a SDR, it has a DAC to generate a certain waveform. The DAC has its noise, which is not at thermal noise level (-173 dBm/Hz), but at higher level. The transmitter RF chain will add its own noise as well.

Definitely, the generated signal is corrupted by noise that is not at thermal noise level.
So, why does the literature perform the calculations by starting from the thermal noise level? It's weird...

Help me to understand where I'm making a mistake.
Thank you

I agree with this consideration in the Wikipedia "Noise figure" article https://en.wikipedia.org/wiki/Noise_figure

noise figure a useful figure of merit for terrestrial systems where the antenna effective temperature is usually near the standard 290 K.

Click to expand...

In other words, it's no useful performance parameter for system that are not noise limited (not receivers or similar impedance matched low level amplifiers for sources with real source impedance), or sources that have a quite different noise temperature.

I didn't still understand :sad:
How to characterize a noise limited system? I've always read about noise figure. All TLC systems I read in my life were characterized with noise figure...

Noise Figure and Equivalent Noise Temperature, both can be used to characterize a receiving system.
Equivalent Noise Temperature is more meaningful and convenient to use when characterize an entire low noise system, when Noise Figure is good to be used to characterize stages, as LNA, Mixers, etc.
How those figures are related you can see in the plot attached:

In fact, "Signal to Noise Ratio" is more important than Noise Figure wherever the signal is.
This will a great insight about the system performance on the overall chain , anywhere.

You want to understand where the thermal noise becomes a problem.
Receivers suffer from thermal noise no matter what the SNR of incoming signals.
In order to assess the receiver performance; first, consider your transmitted signal has perfect SNR, assume it even noiseless. Actually, in reality, you do not worry about the SNR of the transmitted signal, its SNR is quite high as long as you provide a moderate transmission.

In the receive side, as the received signal level becomes lower and lower, you will encounter the thermal noise as dominating noise source restricting the receive operation.

Yes, the thermal noise set the limits in a very low noise receiving system.
This is the reason that the concept of Equivalent Input Noise Temperature is used in low noise satellite and space receiving systems.