RIS phase optimization with coverage map

[Sina7474]

Hello,

I am using Sionna ray tracing to compute coverage maps while deploying RIS. In some cases, I need to optimize the RIS phase profile, but I cannot use the ris.phase_gradient_reflector(source, target) function. I attempted to implement RIS phase profile optimization by computing path gain through the coverage map, but the optimization did not work in this case. However, when I use Path.CIR for computation, the optimization framework functions correctly. Interestingly, if I configure the RIS phase using ris.phase_gradient_reflector(source, target), the coverage map works fine, but the optimization method still fails.

I adapted your multi-mode RIS optimization example to a simplified version with a single RIS, one receiver (tile), and one Tx. I am not performing joint optimization; instead, I aim to optimize the RIS phase values solely based on the path gain computed from the coverage map function.

I have attached a code where the RIS phase profile fails to optimize using the coverage map. I would greatly appreciate your help in identifying the issue.

NLoS_optimization_test.zip