How can I design flat magnitude response IIR filter ?

I want to design flat magnitude response IIR filter to simulate the memory effect of High Power Amplifier nonlinearities

Answer 1

[Star Strider]

I am not certain what you want the filter to do. One option is to design a lowpass filter with a cutoff near the Nyquist frequency. There are several ways to implement such a filter, and filter your signal.

It would help if you describe ‘the memory effect of High Power Amplifier nonlinearities’ in more detail. I am familiar with AF and RF power amplifiers, however I am not aware of this effect.

[Abdelwahab Afifi]

There are two effects of the Power amplifier :

1- Nonlinearities

where the output is afunction in the several orders of the input sample

2- Memory effect

where the output depend not only on the current input samples but also the previous samples

The nonlinearities can be simulated using either SALEH model or RAPP model in simulink. In the following paper, the author try to simuate the memory effect using Asymmetric IIR filter. I read about IIR filter that ( IIR filters introduce frequency-dependent delay ). . But I'm not sure why he use asymmetric filter?! I think we need IIR filter to provide delay for the signals / not distortiong the amplitude. any recommendations about simulating a complete PA model with integrate the two effects using MATLAB?

https://www.mathworks.com/company/newsletters/articles/adaptive-dpd-design-a-top-down-workflow.html

[Star Strider]

Unfortunately, I cannot help you with this. I have not used Simulink in a while, and I cannot design filters with it.

From your initial description, I thought that this was a much more straightforward problem than it turned out to be.

[Abdelwahab Afifi]

when I tried to use the given parameters to simulate the filter response, I get a different response? can you help me to estimate the parameters that give the original response ?