Multimodal direct-ray transmission for enhanced point-to-point wireless-channel capacity
Phys. Rev. Applied 26, 024030 – Published 12 August, 2026
DOI: https://doi.org/10.1103/ts68-vtpl
Abstract
Line-of-sight point-to-point links are typically rank-limited when they are driven and observed only through conventional uniform-aperture far-field ports: with finite apertures and no scatterers, the dominant contribution is the Friis mode, whose received power scales as . We introduce Multimodal Direct-Ray Transmission (MDRT), a mode-synthesis and link-budget framework derived from an asymptotic expansion of the free-space Green’s function that retains radiative near-field terms. The expansion identifies deterministic coaxial current distributions whose field amplitudes scale as (power ), with recovering the classical Friis term and producing first-order secondary channels with power . MDRT does not create degrees of freedom beyond the full dyadic Green-function channel; rather, it provides closed-form current weights, scaling laws, and generalized Friis-like formulas for accessing weak but existing higher-order singular channels. Because the relevant weights reduce to uniform, linear-gradient, and monopulse profiles, the leading MDRT channels can be realized as a small number of passive or hybrid modal antenna ports, reducing the rf-chain and calibration burden relative to fully digital element-level multiple input multiple output and singular value decomposition (SVD) implementations. We extend the analysis to the dyadic Green’s function and express the first higher-order link budget in terms of a dimensionless first-order tensor directivity, interpretable as the squared angular slope of the normalized far-field vector at a boresight null. Numerical operator SVD confirms that the predicted current distributions approximate the dominant higher-order singular modes in the Fresnel or transition region. Finite-signal-to-noise ratio (SNR) capacity tables quantify when these modes are above the water-filling threshold for realistic frequencies, apertures, ranges, and receiver SNRs. Finally, a two-mode prototype demonstrates modal separation and the measured versus decay. The method requires accurate boresight alignment and is most relevant near the far-field boundary of large electrical apertures.