Generative Neural Networks for Simulating Radio Emission from Air Showers

The simulations of radio emission from EAS are essential for reconstructing various shower param-eters from the measured radio signals. As bigger experiments use more and more antennas, thecomputational cost of these simulations gets prohibitively large. These simulations also scale ex-ponentially with higher primary energies and linearly with the number of antennas. Thus thereis a need to interpolate and generate radio signals across various energy ranges and antenna posi-tions.In this work, we present a novel neural network which can predict radio pulses for the AERA setupusing several shower parameters and antenna positions as input. The results which showcase thepulses generated by the network compared to the CoREAS simulations, are presented. The network’sability to also get the fluence pattern and the total radiation energy is shown along with its perfor-mance benchmarks. Finally, the network is used in a simplistic Xmax reconstruction procedure toshow the viability of these generated pulses for Xmax reconstruction.