Drifting Time-Navigation Beacons (DTNB): A Solar-System GNSS Architecture with Built-In Residual Observatory — Protocol Specification, Synchronization Mesh Fitting, and Observational Consistency Analysis

We propose a Solar-System GNSS architecture based on drifting, long-lived Time-Navigation Beacons (DTNBs) deployed at strategically chosen heliocentric locations (e.g., Earth-Sun L1/L2, Mars vicinity, main belt). Each beacon continuously broadcasts signed clock state, ephemeris, and integrity metadata, enabling spacecraft to autonomously correct onboard time and position with reduced dependence on Earth-based tracking. Crucially, DTNB operations produce continuous, standardized residual logs by design, enabling multi-point correlation and closure observables suitable for Long Time-Base Interferometry (LTBI)-style detection of low-frequency synchronization fluctuations. We define a minimal protocol/API with two-mode operation (NAV vs SCI), present falsifiable performance metrics, and outline deployment roadmaps and governance options. As a concrete science use case, we introduce Synchronization Mesh Fitting — a method that exploits the three-dimensional geometry of the DTNB constellation to estimate field parameters of spacetime residual correlations at precision levels significantly exceeding individual clock accuracy (10⁻¹⁹ from 10⁻¹⁶ USOs via structured 3D fitting). Preliminary analysis of existing GPS and deep-space probe clock data provides suggestive evidence that Ornstein-Uhlenbeck-like correlation structure may exist in timing residuals across 0.36–20 AU; if confirmed, DTNB is designed to characterize this structure systematically. We demonstrate that the hypothesized signal is a scalar (spin-0) breathing mode — experimentally distinguishable from tensor gravitational wave backgrounds via its angular correlation signature (exponential baseline decay vs. Hellings-Downs quadrupole). The signal is spectrally suppressed at frequencies probed by existing Lorentz invariance tests (>5 orders below optical cavity sensitivity at 10³ Hz), and is absorbed as common-mode noise by current PTA pipelines, explaining the absence of prior detection. The falsifiable claim is that the spatial correlation length of post-fit timing residuals is non-zero (ℓ_c > 0), whereas general relativity combined with standard instrumental noise models predicts ℓ_c = 0. The proposed system transforms deep-space navigation services into a persistent observatory as an operational byproduct, independent of whether this specific correlation structure is ultimately validated. This work is part of the YAGC (Yoshida-AI-Generated Cosmology) Project, a collaborative theoretical physics initiative employing systematic dialogue with multiple AI systems.