Hi RF designers,

In most of the publications the RF transistors of an LNA are almost NMOS transistors only. If we assume that PMOS devices are fast enough for the operating frequency, would it make sense to employ the PMOS as well?

Thinking in the frame of one stage designs, and assuming that a pmos is a similar transistor than the nmos, we have two devices connected in parallel. So we might have twice the gm but twice the gds as well for the same current, so neither the gain, nor the output noise change.
Here I have assumed that the real part of the output impedance is determined by the output impedance of the transistors. Is it true in general?

For me it is a strange result, that having a pmos as well does not improve the circuit performance.  Could you confirm or debunk the above statement? Can you give examples where using both N and PMOS improves the performance?

Adding to what I wrote above, a load is usually needed on top of NMOS. If that load does not contribute signal gain and is simpy a current source, then it will add noise. Hoever, if the load takes part in signal amplification, then it might still contribute the same noise but it also contributes to the signal amplification thus lowering NF.