TAs civil aviation and 5G technologies continue to advance, there is an increasing demand for stable, high-speed in-flight internet connectivity. Traditional handover algorithms, designed for ground-based networks, struggle to meet the demands of high-speed mobile environments, such as those experienced in air-to-ground (A2G) communication systems. This paper introduces a handover strategy tailored for A2G scenarios, leveraging a fast prediction model to adapt to rapid signal fluctuations caused by aircraft movement. The proposed strategy dynamically adjusts handover decisions based on flight speed and course angle, aiming to optimize the handover frequency and success rate. The prediction model preprocesses received signal strength data to reduce variance in handover measurements, while adaptive hysteresis threshold adjustment based on aircraft speed minimizes unnecessary handovers. Additionally, by incorporating course angle information, the strategy restricts handovers to cells aligned with the aircraft’s trajectory, reducing redundant handovers and enhancing overall system stability. Experimental results demonstrate that this approach significantly improves connectivity continuity, reduces handover failures, and lowers system load, providing a more reliable and efficient in-flight connectivity solution. This adaptive handover strategy presents a promising approach for enhancing A2G communication performance, paving the way for more robust, user-friendly in-flight network experiences in next-generation aviation communications.