Abstract:

The simulation of acoustic and electromagnetic wave propagation is essential for understanding environmental impacts, advancing wireless communication, and optimizing architectural and sensing systems. Open-source simulators provide reproducible, accessible, and extensible platforms that can be readily integrated with machine learning (ML) frameworks. This integration enables data-driven optimization, inverse problem solving, and synthetic dataset generation, which are increasingly important for both scientific research and business applications such as smart city planning, environmental monitoring, and next-generation communication networks. This paper surveys state-of-the-art open-source simulation environments, focusing on their programming foundations, dimensional capabilities, and support for ML integration. The evaluation considers both advantages and limitations, with emphasis on computational requirements, simulation fidelity, scalability to large-scale problems, and integration with ML.