Lookahead‐Based Lookaside Selection: Mitigating Inline Frequency Interference Between Satellite Constellations

ABSTRACT The densification of non‐geostationary orbit (NGSO) constellations has made co‐frequency interference (CFI) a critical bottleneck for system capacity, particularly inline interference events where angular separation between heterogeneous satellite systems approaches zero. Existing mitigation strategies, such as spatial isolation or reactive handovers, often suffer from myopic decision‐making, resulting in either persistent interference or excessive signaling overhead. This paper proposes a proactive lookahead‐based lookaside selection algorithm to mitigate inline CFI in dynamic NGSO environments. By modeling satellite selection as a receding horizon optimization problem, the proposed algorithm integrates multi‐step trajectory prediction, a temporal discount factor, and explicit switching penalties to optimize the long‐term radio frequency environment. Simulation results using Walker‐type constellations over a 30‐day period demonstrate that the proposed method effectively eliminates severe interference events. At high latitudes (60° N), the probability of angular separation falling below 5° is reduced to 0%, compared to 1.1% for conventional highest‐elevation strategies. Similar robustness is observed at lower latitudes and in challenging co‐inclined “chasing” scenarios. Crucially, this performance is achieved with a balanced operational cost: Daily handovers increase by only 64% relative to stability‐first baselines, while remaining 70% lower than elevation‐maximization strategies. Parametric analysis further reveals that the optimal lookahead horizon correlates strongly with the average satellite pass duration, establishing a design guideline for proactive interference management in future 5G/6G‐integrated non‐terrestrial networks.