Active device nonlinear modeling

What is the difference in nonlinear modeling using X-parameter and conventional device modeling using analytical approach.

Conventional modelling takes a standard model into account and defines model parameters by measurements/computations.
X-Parameters are behavioural nonlinear representation of linear/nonlinear devices and they are not related to any certain models.They consist of mathematical equations that emulate the nonlinear behaviour
of active devices.They can also be defined measurments and those equations are defined by curve fitting into linear/nonlinear mathematical expressions.

Practically when we are testing GaAs FET device ( depletion mode) , if the gate looses its control, then suddenly gate current increases and also the drain current increases abruptly, but we donot observe any RF gain? e

Why the device noise temperature rise with frequency?

If RF/LO isolation is very poor in FET based mixer ( where LO modulates the transconductance), what will happen? Here LO and Rf both applied simultaneously to the gate of the FET through their filters.

SOME BASIC POINTS I WANT TO NEED REGARDING NONLINEAR MODELING of MESFET/HEMT:

a. DC I-V characteristics of the DEVICE IS SUFFICIENT ? to characterize the device alone !

b. what is the role of s-parameters in nonlinear modeling? Why S-parameters at various bias points are required?

- - - Updated - - -

Dispersion is a low frequency phenomena but some of the commercially available models include the dispersion effect! i am trying to extract model nonlinear parameters of a HEMT device ( from 1GHz-to 18GHz) , how the performance predictions for the large signal model will change if we discard the dispersion effect!