Electromagnetic Information Analysis for Joint Communication and Sensing in Complex Scattering Environments

This study addresses a limitation in Shannon information theory by characterizing electromagnetic (EM) information metrics for joint communication and environmental sensing in complex scattering environments. We specifically characterize how uncertainty introduced by a dynamic scattering environment impacts these metrics, departing from classical methods that treat additive white Gaussian noise (AWGN) as the primary uncertainty source at the receiver. Our analysis examines EM information transfer within an environment containing multiple scatterers whose positions rapidly change causing fast fluctuations in channel characteristics. The proposed framework eliminates the need to assume ideal channel coherence times for deterministic estimation or use cumbersome channel state forecasting. Mutual information metrics are derived using conditional probability distributions to quantify both information transfer between source and receiver and inference of environment properties (characterized by the scatterer density in this study) at the receiver. The scattering environment is modeled with randomly distributed spherical dielectric scatterers positioned between a pair of static dipole antennas. Scattered fields are computed using the Foldy-Lax multiple scattering theory (MST) in Monte Carlo electromagnetic simulations, with key results compared against single-scattering approximation (SSA) counterparts to illustrate the significance of incorporating comprehensive scattering physics in information modeling. Source symbols encoded via M-ary amplitude shift keying (M-ASK) and Mary quadrature amplitude modulation (M-QAM) are processed non-parametrically at receiver by estimating conditional probabilities from E-fields data using kernel density estimation (KDE).