S11 and Frequency Modeling for Physical Parameter

AI Thread Summary
The discussion revolves around creating an empirical model that relates frequency and S11 (reflection coefficient) to a physical parameter, specifically electrical properties of a material. The user has conducted simulations using CST and recorded S11 data from 8-15 GHz but is struggling to derive a model that can predict electrical properties from S11 measurements. There is confusion regarding the specifics of the physical parameter being targeted and the inputs required for the model. It is suggested that the user needs to clarify their objectives and consider the concept of inverse problems to effectively reverse-engineer the relationship between S11 and the desired properties. The conversation emphasizes the importance of specificity in defining the problem to obtain useful guidance.
labd0001
Messages
9
Reaction score
0
Hi all,

I am trying to obtain an empirical equation/model having 2 inputs (frequency and S11) and one output (a physical parameter)
I have done various simulations using CST of the System and now I am trying to relate S11 and frequency to a physical parameter. Any suggestions of how I should go about it?

I have tried using system identification tool in Matlab but the results I obtained are rather confusing and I am not sure whether the approach is correct.

Thanks
 
Engineering news on Phys.org
You have described a box with three wires (I'm guessing it's electronic, it could be three tubes or three nerves or something - you havn't actually said) coming out of it.
Two are labelled "input" and one "output" ... the input wires are further labelled "frequency" and "S11".
There is no description of what is in the box or what these labels actually refer to.

You want the "output" to be something to do with a physical parameter" but you don't know which parameter and you have not said what would count - presumably the signal that comes from the output is not, itself, a physical parameter.

Basically the request is too wide open for any answer to be useful.
Usually, the act of describing your problem is sufficient detail and specificity to get a useful answer will result in an answer occurring to you.
 
I would like to relate (empirically) the reflection coefficient of a material with a physical property being length, electrical properties etc for specific frequencies. That is why the input of the model/equation will be the frequency and reflection coefficient (at that particular frequency) and the output is the physical property I am relating to.
 
OK - you you want your box to tell you some length given the reflection coefficent and frequency?
You have said that the reflection coefficient is for a material - presumably the material surface?
What is the frequency of: incoming light perhaps? You have to say. The software you are using is a tad general.
Finally, "length" of what?

I don't know what " electrical properties etc for specific frequencies" means in this context - please be specific.
 
frequency of incident signal. I am doing a sweep from 8-15GHz and recording the S11.
What i did is the following, I simulated a material (with known electrical properties) and calculated the S11 for 8-15GHz. Therefore I have the S11 for particular frequencies and their corresponding electrical properties. Now I would like to obtain a model so if I do measurements of S11 in the laboratory using the same system I used in the simulation, I can calculate the electrical properties.

I hope this makes it clearer.
 
It would be clearer to me if I knew what a 511 is.
 
S11 is the reflection coefficient measured using a vector network analyser
 
OK - you have a computer model already that takes some "known electrical properties" (whatever that means) and gives you the reflection coefficient as a function of incident frequency.

But what you want another model that goes the other way ... takes the data for S11(f) and tells you the "electrical properties".

It is unlikely that this is possible - in general, the reverse mapping won't be unique.

Inside the computer model is a set of calculations ... to get the reverse model, you need to reverse the calculations.
For more information on how to get a model (in your case, the "electrical properties") from data, have a look at "inverse problems" from S.M.Tan, here: http://home.comcast.net/~szemengtan/ ... that should give you an idea of how much vital information you have left out.

Until you are more forthcoming I cannot help you.
Good luck.
 
Last edited by a moderator:
Labd0001, it would help if you told us exactly what you are trying to do. Could you go into more detail please? What exactly are you trying to do?
 
Back
Top