Research on predictive handover mechanism for low-altitude heterogeneous networks based on multisource situation fusion

In low-altitude communication environments, factors such as spatial dynamics, link instability, and coexisting heterogeneous networks render traditional instantaneous threshold-based handover strategies incapable of ensuring service continuity. To enhance network access stability in low-altitude scenarios, this study proposes a predictive handover mechanism oriented toward multi-source situational awareness. Methodologically, this approach unifies encoding of multi-source information—including link quality, flight status, and environmental obstructions—to generate future link evolution trajectories via a temporal prediction model. Dynamic trigger thresholds are constructed based on the prediction sequence, with time-weighted risk functions identifying degradation phases. Building upon this foundation, a heterogeneous network selection strategy is designed, employing a utility function for comprehensive evaluation of candidate networks. The experimental system integrates the ns-3 heterogeneous network simulation environment with real flight trajectories, validating performance across three scenarios: urban blocks, elevated corridors, and semi-open areas. Results demonstrate that this mechanism outperforms existing strategies in handover success rate, latency stability, and invalid handover control, with significantly improved trigger lead time and consistent decision behavior across scenarios. The study indicates that proactive handover methods based on situational fusion and future link inference can effectively enhance service continuity in low-altitude heterogeneous networks, providing insights for intelligent management of future large-scale airspace communications.