About thecentral frequancy

[HT.T]

Dear Craig

  

      When I analyze the spectrum of the direct wave, I find that the frequency of the spectrum center, that is, the center frequency is different from the detection frequency. For example, 

when the detection frequency is 900 MHz, the spectrum analysis of the direct wave shows that the center frequency is about 1000 MHz, and when the distance between the receiving 

and transmitting antennas is very close, it is even higher than 1000 MHz. What is the reason for this。

 

                                                                                                                                                                                                                                                                                       HT.T

this is my code

#domain: 0.6 0.6 0.002 

#dx_dy_dz: 0.002 0.002 0.002

#time_window: 40e-9

#waveform: ricker 10 1200e6 my_ricker

#hertzian_dipole: z 0.3 0.30 0.001 my_ricker

#rx: 0.9 0.30 0.001

#src_steps: 0.01 0 0

#rx_steps: -0.01 0 0

#geometry_view: 0 0 0 1.2 0.6 0.002 0.002 0.002 0.002  free_space  n

[Antonis Giannopoulos]

Your inout file has an error in the #rx: command but even if you fix this you will get discrepancies in the frequency analysis. There are few reasons that you get this. 

The first and most important one has to do with the actual electromagnetic model. When solving a 2D problem you are using a Line Source which is infinite long. The impulse response of this source is not finite and this is what is causing the different effect in the frequency of the response as you see it. Your result is the convolution of this solution with your excitation and hence is affected by the nature of the impulse response in this way. You can test this with the analytical solution as it will agree with gprMax. However, you need to be careful as in 2D is the solution of a line source and not a dipole source.

In gprMax when you are doing a 2D simulation and you are using a #hertzian_dipole: you are actually using a line source. If you are doing a 3D simulation you are using a dipole source.

If you do this experiment in 3D you will see that the frequency is not changing (if no other errors are there) as the impulse response (i.e. the Green's function) is of a finite duration. 

The other error that you need to be careful off (not so much here but it is something to remember) is to have appropriate frequency resolution. For 40 nanoseconds time window your frequency resolution is 25MHz so your FFT bins are a bit broad and you can be +/- around there in determining your peak frequency.

Hope that helps 

Antonis