地球静止轨道（GEO）卫星链路受低轨（LEO）通信星座干扰的问题日趋严峻，在LEO星座设计阶段考虑链路干扰影响可以帮助缓解这种问题。提出了一种考虑链路干扰的低轨通信星座优化设计方法，改进了干扰规避区域，并建立干扰评估的半解析模型提高了计算效率。首先，以星座轨道及构型参数为设计变量，以LEO星座建设成本及其对GEO卫星链路干扰水平为目标函数，考虑通信范围、地面站跟踪策略、星座不间断通信、构型限制等约束，建立了LEO星座优化设计问题模型；然后，设计了改进干扰规避区域，并利用低轨卫星通过干扰规避区域的时间占比来评估星座干扰水平；最后，采用非支配排序遗传算法进行优化求解。结果表明，改进的干扰规避区域可以更有效地表征不同轨道高度、不同地面站分布下的有害干扰区域分布；设计的时间占比指标与传统的干扰评价指标相似性达0.998，且计算速度提升了约40倍；卫星相位优化能够在不新增成本的情况下使GEO下行链路受到的干扰水平降低超过5%，星座构型优化能够在新增成本小于原成本1/3的情况下使干扰水平降低超过90%。

The issue of the interference from the Low Earth orbit (LEO) communication constellation to the Geostationary Earth orbit (GEO) satellite links has become increasingly severe. Considering the impact of link interference during the design phase of LEO constellation can help mitigate this issue. This paper proposes an optimized design method for LEO communication constellation that incorporates link interference consideration, which designs an improved interference exclusion zone and establishes a semi-analytical interference assessment model to enhance computa-tional efficiency. First, a multi-objective optimization model for LEO constellation design is established, with the con-stellation orbital and configuration parameters as design variables, and the LEO constellation construction cost and interference levels on GEO satellite links as objectives. Constraints include the communication coverage, ground station tracking strategy, uninterrupted communication of constellation, and configuration limitations. Second, an improved interference exclusion zone is designed, and the interference level of the constellation is evaluated using the time proportion of LEO satellites passing through this zone. Finally, the non-dominated sorting genetic algorithm is employed to solve the optimization problem. Results demonstrate that the improved interference exclusion zone can more effectively characterize harmful interference zones under different orbital altitudes and ground station dis-tributions. The proposed time proportion metric exhibits a similarity of 0.998 with the traditional interference evalua-tion index while achieving a 40-fold acceleration in the computational efficiency. The optimization of the satellite phase reduces the interference level on GEO downlinks by over 5% without the additional cost, and the optimization of the constellation configuration reduces the interference level by more than 90% with a cost increase of less than one-third of the original design.