Virtual Extremely Large Antenna Arrays: Emulating Near-Field Wideband Beamforming in Dense Cell-Free Networks

Cell-free (CF) massive MIMO (m-MIMO) architectures operating under Rician fading conditions are investigated, with a focus on position-aware signal processing and Bayesian channel estimation. By treating distributed access points (APs) as components of a virtual extremely large aperture array (V-ELAA), the paper leverages near-field (NF) propagation and geometric channel characteristics of CF systems and highlights their potential to deliver fine-grained spatial resolution and improved communication and localization performance. We propose a Cramér-Rao-based analysis to quantify localization accuracy, revealing that accurate user positioning enhances channel estimation performance, especially as the channel becomes increasingly deterministic due to a dominant line-of-sight (LoS) component. Numerical evaluations of effective degrees of freedom (EDoF) across different deployment densities demonstrate the spatial diversity advantage of the CF architectures with respect to co-located m-MIMO deployments. Results show that the considered CF m-MIMO network deployment may approach perfect CSI performance through accurate localization. Finally, scalable user-centric CF implementations are examined, highlighting favorable tradeoffs between computational complexity and achievable SE.