WITHDRAWN: SLAM-Enabled Localization and GNSS Recovery Strategies for Autonomous Spacecraft Navigation

Future satellite constellations, deep-space missions, and cislunar operations face significant challenges in navigation when Global Navigation Satellite System (GNSS) signals are weak, unavailable, or deliberately denied. Conventional inertial navigation and visual odometry approaches provide only short-term accuracy and suffer from drift during extended operations. This paper presents a SLAM-enabled framework for drift-free localization, integrated with GNSS recovery mechanisms, to enhance autonomous spacecraft navigation and resilience in contested or GNSS-denied orbital environments. The proposed system leverages monocular visual SLAM to map orbital or planetary surroundings while simultaneously estimating spacecraft position. Loop closure detection mitigates long-term drift throughout extended GNSS outages. Visual-Inertial Odometry (VIO) integrates inertial measurements with visual cues for real-time state estimation. When GNSS signals are reacquired, a sigmoid-weighted recovery mechanism seamlessly realigns the SLAM/VIO trajectory with the GNSS solution, eliminating trajectory discontinuities. Simulation studies using synthetic orbital imagery and IMU/GNSS datasets demonstrate that the SLAM-based system reduces position RMSE by 50–72% over standalone VIO across 60-, 300-, and 900-second outage scenarios, with the sigmoid GNSS recovery filter reducing maximum position rate-of-change from 878.87 m/s (hard switch) to 95.23 m/s.