With the advent of the mobile Internet of Things (IoT) era, smart devices have become deeply integrated into our daily lives, leading to an exponential increase in spatial data generated daily. Outsourcing such data to cloud servers has become a widely adopted solution, but it brings along concerns of data breach and security. The existing solutions lack the capability for efficient retrieval of encrypted data while also incurring substantial storage overhead. To address the aforementioned challenges, we propose a spatial data indexing structure leveraging the Geohash algorithm, which integrates a prefix-filled tree and a Bloom filter with corresponding efficient pruning mechanisms. Specifically, we utilize the Geohash algorithm to construct a spatial position index tree, filling in the spatial position index vectors for each node to achieve rapid matching, and use Bloom filters to address key attribute storage of large-scale spatial data. This design significantly reduces the number of spatial matching operations while minimizing storage overhead, thereby improving overall search efficiency and achieving sub-linear retrieval performance. Comprehensive evaluations on real-world datasets reveal that our approach outperforms existing methods, offering roughly 8 times faster query efficiency and a 119% increase in storage efficiency.