Radiation Pattern Prediction for Metasurfaces: A Neural Network-Based Approach

As the current standardization for 5G networks nears completion, work towards understanding potential technologies 6G wireless is already underway. One of these reconfigurable intelligent surfaces. They offer unprecedented degrees freedom engineering channel, i.e., ability to modify characteristics channel whenever and however required. Nevertheless, such properties demand that response associated metasurface well understood under all possible operational conditions. While an radiation pattern can be obtained through either analytical models or full-wave simulations, they suffer from inaccuracy extremely high computational complexity, respectively. Hence, in this paper, we propose a neural network-based approach enables fast accurate characterization response. We analyze multiple scenarios demonstrate capabilities utility proposed methodology. Concretely, show method learn predict parameters governing reflected wave with accuracy simulation (98.8–99.8%) time complexity model. The aforementioned result methodology will specific importance design, fault tolerance, maintenance thousands surfaces deployed network environment.