Free-space prapagation function using Huygens spherical convolution

Given a certain field information at input plane, we can predict any field of view at output plane

Given a certain field information at g1 plane, we can predict any field of view by defining x2half and dx2 (assuming the field is only polarized along x or y direction)

Advantage:
Great for fine and small g1 and relatively large g2 plane matrix. Also, free from replica generated from Fourier Transform approach. Allow different sampling rate at g1 and g2 planes. Good for analyzing finely sampleddata generated from FDTD software near field
Disadvantage:
Very slow if g1 is a large matrix. Recommend size of g1 < 500x500
g1: Field information at initial plane
dx1: Grid size at g1 plane
dx2: Grid size at g2 plane
Nx2: Number of pixels at the target plane (assuming centered)
z: Propagation distance
lambda: Wavelength used

Example:

g1=zeros(60);g1(30,30)=1;
dx1=10e-6; dx2=1e-6; z=0.01;lambda =532e-9; x2half=1e-4;
[g2]=ConvProp(g1,dx1,dx2,z,x2half,lambda)
imagesc(abs(g2))

NOTE: x2half is the distance from the end of camera to the center subtracted by half pixel length

eg: pixel size = 1 um & 5 pixels.
|1|2|3|4|5|
x2half = 5um/2 - 1um/2 = 2 um

---Disi A Sep, 6th, 2013
[e-mail]