• State TI Thinks Resolved
  • Locked Locked
  • Replies 2 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 823 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

IWR1642BOOST: The symmetry in range FFT result from mmStudio PostProc

user5888640
Prodigy 20 points
Part Number: IWR1642BOOST
Other Parts Discussed in Thread: IWR1642

Hi guys,

I'm using IWR1642  to detect obstacles. The mmStudio PostProc can generate 1D-FFT amplitude profile with complex, only-real and only-imag modes. The results are shown below.

My question is: why does the symmetry of FFT preserve both in the only-real and only-imag modes, but does not preserve for the complex case?

Since mmStudio is not open-source, when implementing my customized range-Doppler FFT, I can only get a mirrored results in the range axis. This could significantly degrade the maximum distance for obstacle detection.

  • 0
    TI__Mastermind 36695 points

    HI, 

    Our mmwave radar has complex receiver chain.  The ADC data is complex value, use only real or imaginary part will loss information.  You should always use complex ADC  data when doing range FFT or Doppler FFT.

    I guess your customized calculation probably only used real or imaginary part of the ADC data.

    It is basic knowledge that a real signal is always symmetric in frequency domain, for example:

    A*cos(wt) = A/2*exp(1j*w*t) + A/2*exp(-1j*w*t);

    However, it is only meaningful to apply FFT on the complex ADC data if you configure the sensor receiver as complex chain.

    Best,

    Zigang

  • 0 in reply to zigang Yang
    Prodigy 20 points

    Thanks, Zigang

    Actually, I indeed did FFT on the complex ADC data.

    I re-examined my Python code for many times and found a severe bug: when realigning the raw ADC samples, I used the reference copy in Python rather than the deep copy!

    After fixing it, I got the correct output:

    Thanks again for the reply!

    Wish you a good day!