Project M03 - Electromagnetic Signatures from Complex Surface Systems and Indoor Scenarios
Future advanced wireless localization and characterization systems – targeted for mm-wave frequencies up to 250 GHz with potentials well into the THz regime – as foreseen for this CRC MARIE – will strongly rely on the accurate retrieval of ultra-broadband electromagnetic signatures from realistic but complex (sub-) surface systems in corresponding indoor scenarios. The central objective of this project M03 is to provide a comprehensive theoretical framework based on hierarchical multi-scale electromagnetic modeling of (randomized) surface morphologies (cf. Fig. 1) that can also be further extended into the macroscopic scales of the proper indoor scenario. Therefore we have to inquire into economic computational electromagnetics based on slim but still accurate numerical engines to represent the characteristic microstructure of the (layered) surface system, and to venture into the very transition from the surface morphology to the indoor environment. Relying here on such an efficient but explicit (multi-scale) fullwave computational electromagnetics approach that follows the required large bandwidth is necessary to grasp the detailed complexity of (sub-) surface morphologies. This reaches way beyond traditional approaches for random rough surfaces such as the Kirchhoff approximation (with its geometrical or physical optics models) and perturbative methodologies, and challenges even the canonical integral-equation formulations of the surface scattering problem as elaborated e.g. in microwave surface scattering  and further qualified in the context of modern amorphous (broadband) nanophotonics . Hence the numerical multi-scale surface model is conceptualized as an ab initio model that may act as a numerical workbench to inquire and test advanced concepts envisaged by our project partners within this CRC. The surplus of this complex surface model will enable us
Regarding the intended hierarchically parametrized multi-scale surface model we will rely on highly economic time-domain simulation engines [F], which – besides our available computational electromagnetics reference codes EMPIRE, COMSOL, FEKO, etc.  – may even be complemented by non-conventional schemes such as e.g. lattice Boltzmann methods [4,5], cellular automata approaches  or stochastic FDTD implementations  that are all apt to cope with large representations of multi-layered randomized surface morphologies. Following the distinct ab initio aspect of our modelling effort allows for in-depth studies of the underlying scattering mechanisms (i.e. surface vs. volume scattering) while inquiring into fundamental questions on the proper representation of random morphologies. It also provides reliable performance estimations for novel material sensing schemes based on MIMO-based reflectometry setups with increasingly compact footprints when looking into near-field and/or bidirectional operation modes (cf. S02) even supported by electromagnetic vortex beams carrying different orbital angular momenta (OAM) as additional degrees of freedom (cf. S03).
The major scientific challenge of M03 may be summarized as follows: to design an efficient parametrized numerical ab initio model of unprecedented detail of the randomized complex surface system that is capable of bridging the gap from physical surface representations to semantic material maps for frequencies up to THz.
In summary, M03 provides a holistic surface model that may act as a numerical workbench for MARIE to forecast/quantify the sensitivity and specificity of the specific material surface characteristics, and to test and inquire advanced MIMO-based reflectometry/ellipsometry setups.