• Part Number SAC2504
• Manufacturer SuperApex Corporation

Try the above mentioned device.

Alan Victor Thank you sir, this low noise transistor does not provide S parameter, noise parameter and equivalent circuit model sir. I need this information's sir, because ADS do not provide Ka band supported transistors directly in their PDKs.

S data is in the data sheet, link below. Contact the company, ask for a Touchstone file and find out if noise parameter data is available.

https://superapexco.com/wp-content/uploads/2019/04/SAC2504-.pdf

Here, try UMS. They have exactly what you need. Its all there.

https://www.ums-rf.com/wp-content/uploads/2017/01/EC2612-99F-Full-1334.pdf

@How can we calculate Linearity measurements by using Touchstone file sir, because s2p file provides linear information only sir. As they provided the equivalent circuit of transistor, but it does not work well. Is there any other way sir?

Correct, the data and model do not include the desired strong signal non linearity. However, the first element of design is low noise figure and adequate transducer gain to minimize second stage contributions degrading SNR. With that said, it is possible to investigate non linearity and deviations from small signal class A operation (LNA typical op is class A) by studying the POWER GAIN and NOISE figure contours created in ADS. Gain compression, limits in Vout and Iout swings from the device load line and the chosen load reflection coefficient will control the large signal performance. An excellent treatment of this is discussed in a Microwaves paper from December 1977. I suggest you visit your university library to obtain the following paper and apply the techniques outlined in: EASY-TO-PLOT GRAPHS SHOW POWER-GAIN TRADEOFFS, PAGES 178-183 author, D. Rosemarin and M. Chorev. Also see... PREDICTING LINEAR POWER AMPLIFIER PERFORMANCE, Microwaves Vol. 13, No 2. pp. 56-59, Feb 1974. Author, K. Richter. These are excellent references bridging linear to non linear when no mon linear model exists. I have written the routines in Mathcad. Easy to do in other tools.

I should add. Both papers address the selection of the load reflection coefficient (LRC) as it directly impacts the output power via the dynamics of the device load line. Either the output power will be current limited or voltage limited. This impacts gain compression, hence device linearity. No mention of a non linear model. Furthermore, the LRC impacts the device input Z and hence the input match which may or may not be near the optimum noise match point. You can see the tradeoff nicely with load power contours and their mapping to the input overlaid on top of input noise contours. Hence, minimizing NF, maximizing linearity, output power and transducer gain are the objectives within your simulation model and setup.