Abstract: The location-based service requires fast query, insertion and deletion operations for research objects, and this demand is augmented for the indoor evacuation fields. Thus, introducing spatial index to tackle the operation efficiency problem is strongly demanded for indoor related scenes, and this method is sounding for indoor spatial objects. Nevertheless, this solution always meets a performance bottleneck. And this performance bottleneck of spatial index pervasively exists in the current compact indoor application scenes. To mitigate this problem, this paper tries to combine the R^* tree index with the Hilbert curve, and proposes an innovative Hilbert curve based index. The succeeding experiment is designed to compare the performance of proposed index with the classic R^* index. The test result shows the new index has successfully alleviated the execution efficiency for multi-type spatial operations, especially on critic indicators of spatial index performance.