APS014 Antenna delay calibration algorithm

Hi all,

I’m trying to implement the algorithm described in the application note aps014. And i can’t seem understand what they are doing in the flowcharts. On page 10, in that flowchart there is following equation:
tof_candidate = (2Del_1 + 2Del_2 + 4*tof_measured)/4
Of what i understand is that the tof_measured is the EDM matrix divided bij speed of light so we get the ToF. And with this formula we make a new matrix, the candidate. The difference between the candidate and the real tof should be close to zero.
Now my questions are:

  • Where do those two delays come from? I suppose from the populate function? And what are they. are they also a 3x3 matrix? If so, the populate function makes 2 3x3 matrices with is 1 candidate?
  • Once we get the candidate that is the best fit. How can the 3 delays be extracted from it. In the example there suddenly appear 3 values, 1 for each device i assume. These values should then be divided by the percentages in table 4 on page 12.

Kind regards,
Bjorn