FPGA Central

Hi everyone,

I am a DSP novice and it is the first time I make a post here. Now, I
am working on a computer program that reads the frequency response from
a vector network analyzer (HP 8753D), and then calculates the time
domain response using inverse FFT. But now I feel that iFFT seems not
enough to get the job done. I have many questions:

1. The network analyzer has a frequency range 30kHz-6GHz, while IFFT
requires evenly-spaced frequency response down to DC. It is a problem
at all to get a correct time-domain response from a frequency response
that is not down to DC?

2. Actually, I think the project is feasible because the network
analyzer has a option that enables it to do time-domain analyses, but
it's just formidably expansive to me. The manual of the network
analyzer said that the time-domain transform is done using something
called "Chirp-Z transform", rather than IFFT that I had expected. Is
Chirp-Z transform the best, or probably the only way to do the trick?

3. If the answer to the previous question is YES, where can I find
some introductory materials about this technique? I have tried to know
more by searching on the web, with no luck.

I appreaciate the replies from you guys.

kawait