Radio wave propagation modeling by combining ray-tracing with integral representations

• Applicant: Professor Dr.-Ing. Thomas Eibert
• Subject Area: Communication Technology and Networks, High-Frequency Technology and Photonic Systems, Signal Processing and Machine Learning for Information Technology
• Term: from 2021 to 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 443636075

Radio wave propagation modelling becomes increasingly important for mobile communications and radar. Mobile communications develops towards techniques such as massive MIMO with very large numbers of antennas. Radar sensing and imaging become ubiquitous in our daily lives. In particular towards the realization of autonomous drive solutions it is of paramount importance to generate accurate radar data in realistic environments by simulation, in order to avoid the collection of radar measurement data in endless drive scenarios. Ray-tracing is clearly the most powerful approach for radio wave propagation modelling in complex environments. However, it has still certain drawbacks, which make it difficult to use at microwave and millimeter wave frequencies. Especially in scenarios, which require several subsequent diffractions, the supposed advantage of high-frequency localization can become a severe drawback, since many (subsequent) diffractions can lead to an enormous amount of computation time and, even worse, the accuracy of multiple diffraction computations is often very bad. The major reason for this accuracy drop is that the localized diffraction computation by evaluating the diffraction coefficients for just one point on an edge is not suitable for the treatment of realistic environments. The applicant has recently established a bidirectional ray-tracing approach combined with the reciprocity theorem, which can already relieve some of the shortcomings of the conventional unidirectional ray-tracing. A key element of this approach is the inclusion of integral representations in order to overcome the stringent localization requirements of the traditional ray-based propagation modelling. This strategy shall be extended in the proposed project, where complex propagation scenarios shall be partitioned based on the Huygens’ and equivalence principles in combination with the already established reciprocity considerations. The individual partitions will be treated by ray-tracing and the interactions between the partitions are obtained based on the corresponding integral representations. In particular for radar scenarios, the procedure will allow to decouple the electromagnetic modelling of the radar targets itself and of the propagation environment. The formulations of the hybrid approach will be worked out and powerful computation algorithms on graphical processing units (GPUs) will be implemented, validated, and optimized, where test and validation scenarios of various complexities will be considered.

• DFG Programme: Research Grants