Stacked Intelligent Metasurfaces: Wavefront Engineering for Extended Effective Radiative Near-Field Range in 6G Systems

Mid- to high-frequency near-field communications and sensing in 6G systems face fundamental limitations when implemented with conventional single-layer programmable metasurfaces. Under fixed physical aperture constraints, these structures suffer from insufficient degrees of freedom for wavefront manipulation, finite phase quantization accuracy, and inadequate capability to synthesize weak-curvature spherical wavefronts at extended distances. Consequently, their engineering-usable near-field range typically falls significantly short of the classical Rayleigh distance. This paper investigates stacked intelligent metasurfaces (SIMs) as a solution to extend the effective radiative near-field range while maintaining identical aperture size, feed configuration, and power constraints. First, we clarify that the Rayleigh distance represents an intrinsic physical boundary determined by aperture size and wavelength, fundamentally constrained by diffraction limits. The proposed "effective range enhancement" does not alter this boundary but reduces the gap between engineering performance and theoretical limits. Second, we establish an equivalent model of SIMs as programmable cascaded lens groups through unified geometrical and physical optics analysis, demonstrating how multilayer structures progressively reshape wavefront curvature via end-to-end joint optimization. Third, we develop a unified near-field model incorporating insertion loss, phase quantization error, and interlayer spacing deviations within a cascaded spherical-wave propagation framework. Numerical simulations validate that multilayer SIMs substantially improve focusing quality and extend the engineering-usable operating range throughout the Rayleigh region, providing theoretical foundation and design guidance for 6G near-field wireless systems.