Configure RIS phase shifts

[FFLMar]

Hello,
I'm currently working on several projects involving the simulation of different RISs inside of a scene. Since among other things, I'm trying to determine wether or not a pre-calculated RIS's phase shift is correct. My question to you is: is it correct if I ONLY change the RIS's phase configuration by imposing it equal to a certain matrix? for example, here is the piece of code I'm using right now:

for i in range(len(SESup)):

    SESup[i] = RIS(name="sesup" + str(i+1),
                position=SESup_pos[i,:],
                num_rows=10,
                num_cols=10, 
                num_modes=1,
                look_at=SESup_pos[i, :] - np.array([0,5,0])) #make it look ahead

    scene.add(SESup[i])

    # Load precomputed phase profile 
    file_name = str(i+1)+"UpPhase_rad_edit.csv"

    phase_profile = np.loadtxt(folder_path + file_name, delimiter=',').reshape(1,10,10)
    phase_profile = tf.convert_to_tensor(phase_profile, dtype=tf.float32) 

    SESup[i].phase_profile.values = phase_profile

# print one of the loaded configs 
SESup[2].phase_profile.show();
plt.title("SES_3Up Phase Profile");

This is the output image of the code above.

Thank you.

[jhoydis]

Hi @FFLMar,

Yes, that should work.

[FFLMar]

After working a bit more on the code I've another question about RISs and their phase configuration.
I managed to compute a phase configuration that is very close to that of Sionna, but if I try to impose it on the ris and run the coverage map function I see nothing. I can't understand why this is the case, the two phase configurations are almost identical.
Here is the code for reference:

n = 0
for i in range (15,16,5):

    scene.add(SES3up[n])

    # Set flat gradient profile
    SES3up[n].phase_gradient_reflector(tx.position, Rx[2].position)
    
    # Show phase profile
    SES3up[n].phase_profile.show(); 
    
    # Compute coverage map with Tx and all SESs
    cm = scene.coverage_map(num_samples=25e6, #increase for better resolution
                        max_depth=4,
                        num_runs=1,
                        los=True,
                        reflection=True,
                        diffraction=True,
                        ris=True,
                        cm_cell_size=[0.01,0.01],
                        cm_size = [6.85, 6],
                        cm_center = [1.6,0.1, 0.9],
                        cm_orientation = [0,0,0])
    
    cm.show(metric='rss', show_ris=True, show_rx = True, vmin=-70, vmax = 0);
           
    scene.remove("ses3up" + str(i) + str(i))    
    n = n+1      

# Initialize new SES3up1515
SES3up[0] = RIS(name="ses3up" + str(15) + str(15),
            position=SESup_pos[2,:],
            num_rows=15,
            num_cols=15, 
            num_modes=1,
            look_at=SESup_pos[2, :] - np.array([0,5,0]))
# Add current SES to the scene
scene.add(SES3up[0])

phase_profile = np.loadtxt("3UpPhase_rad_edit_sionna.csv", delimiter=',').reshape(1,15,15)
phase_profile = tf.convert_to_tensor(phase_profile, dtype=tf.float32) 
   
# Set loaded config
SES3up[0].phase_profile.values = phase_profile

# Show phase profile
SES3up[0].phase_profile.show(); 

# Compute CM
cm = scene.coverage_map(num_samples=25e6, #increase for better resolution
                        max_depth=4,
                        num_runs=1,
                        los=True,
                        reflection=True,
                        diffraction=True,
                        ris=True,
                        cm_cell_size=[0.01,0.01],
                        cm_size = [6.85, 6],
                        cm_center = [1.6,0.1,0.9],
                        cm_orientation = [0,0,0])

# Show CM
cm.show(metric='rss', show_ris=True, show_rx = True, vmin=-70, vmax = 0);

As you can see the tow phase profiles are not dissimilar. I would expect to see at least something in the CM.
I've also changed the coverage map z to see if I was "cutting" it at the wrong point, but nothing changed.